Yu Zhang

All Publications

• Brain-Cognition Fingerprinting via Graph-GCCA with Contrastive Learning. Machine learning in clinical neuroimaging : 7th international workshop, MLCN 2024, held in conjunction with MICCAI 2024, Marrakesh, Morocco, October 10, 2024, proceedings. MLCN (Workshop) (7th : 2024 : Marrakesh, Morocco) Wang, Y., Peng, W., Zhang, Y., Adeli, E., Zhao, Q., Pohl, K. M. 2025; 15266: 24-34

  Abstract

  Many longitudinal neuroimaging studies aim to improve the understanding of brain aging and diseases by studying the dynamic interactions between brain function and cognition. Doing so requires accurate encoding of their multidimensional relationship while accounting for individual variability over time. For this purpose, we propose an unsupervised learning model (called Contrastive Learning-based Graph Generalized Canonical Correlation Analysis (CoGraCa)) that encodes their relationship via Graph Attention Networks and generalized Canonical Correlational Analysis. To create brain-cognition fingerprints reflecting unique neural and cognitive phenotype of each person, the model also relies on individualized and multimodal contrastive learning. We apply CoGraCa to longitudinal dataset of healthy individuals consisting of resting-state functional MRI and cognitive measures acquired at multiple visits for each participant. The generated fingerprints effectively capture significant individual differences and outperform current single-modal and CCA-based multimodal models in identifying sex and age. More importantly, our encoding provides interpretable interactions between those two modalities.

  View details for DOI 10.1007/978-3-031-78761-4_3

  View details for PubMedID 39872150

  View details for PubMedCentralID PMC11772010

• Multiband EEG signatures decoded using machine learning for predicting rTMS treatment response in MDD. Journal of affective disorders Arteaga, A., Tong, X., Zhao, K., Carlisle, N. B., Oathes, D. J., Fonzo, G. A., Keller, C. J., Zhang, Y. 2025: 119483

  Abstract

  Repetitive transcranial magnetic stimulation (rTMS) is a promising treatment for major depression disorder (MDD), particularly for treatment-resistant cases. However, identifying translatable biomarkers predictive of treatment outcomes remains underexplored.Participants with treatment resistant depression from the TDBRAIN dataset underwent either high frequency rTMS (10 Hz) at the left dorsolateral prefrontal cortex (DLPFC) (Protocol 1, n = 44) or low frequency rTMS (1 Hz) at the right DLPFC (Protocol 2, n = 73). Pre-treatment electroencephalograms (EEG) was collected, and changes in Beck Depression Inventory were measured post-treatment. EEG oscillations were decomposed into multiband intrinsic mode functions (IMF) and integrated under a latent space predictive modeling framework to identify signatures for predicting treatment outcomes.Multiband signatures significantly predicted rTMS outcomes (Protocol 1: r = 0.40, p < 0.01; Protocol 2: r = 0.26, p < 0.05). Key spatial patterns linked to treatment outcomes were identified, revealing three main oscillations: IMF-Alpha, IMF-Beta, and the residual signal. In Protocol 1, critical regions included the left frontal and parietal regions for IMF-Alpha, left frontal-central and right parietal regions for IMF-Beta, and bi-hemispheric central and left parietal-occipital regions for residual signals. In Protocol 2, critical regions involved the left frontal and parietal regions for IMF-Alpha, left frontal-central region IMF-Beta, and right frontal, left frontal-central, midline central, and left parietal-occipital regions for residual signals. These oscillatory features also showed correlations with specific personality measures, suggesting their potential clinical relevance.Our findings demonstrate the promise of machine learning-driven multiband EEG signatures for personalized MDD treatment prediction, offering a translatable pathway for improved patient outcomes.

  View details for DOI 10.1016/j.jad.2025.119483

  View details for PubMedID 40441660

• Contrastive functional connectivity defines neurophysiology-informed symptom dimensions in major depression. Cell reports. Medicine Zhu, H., Tong, X., Carlisle, N. B., Xie, H., Keller, C. J., Oathes, D. J., Liu, F., Nemeroff, C. B., Fonzo, G. A., Zhang, Y. 2025: 102151

  Abstract

  Major depressive disorder (MDD) is highly heterogeneous, posing challenges for effective treatment due to complex interactions between clinical symptoms and neurobiological features. To address this, we apply contrastive principal-component analysis to fMRI-based resting-state functional connectivity, isolating disorder-specific variations by contrasting data from 233 MDD patients and 285 healthy controls. Subsequently, we use sparse canonical correlation analysis to identify two significant dimensions linking distinct brain circuits with clinical profiles. One dimension relates to an internalizing-externalizing symptom spectrum involving visual and limbic networks and is associated with cognitive task reaction times. The other dimension, linked to personality traits protective against depression (e.g., extraversion), is driven by dorsal attention network connections and correlates with cognitive control and psychomotor performance. This approach illuminates stable symptom dimensions and their neurophysiological underpinnings, aiding in precision phenotyping for MDD and supporting the development of targeted, individualized therapeutic strategies for mental health care.

  View details for DOI 10.1016/j.xcrm.2025.102151

  View details for PubMedID 40441140

• Multi-modal cross-domain self-supervised pre-training for fMRI and EEG fusion. Neural networks : the official journal of the International Neural Network Society Wei, X., Zhao, K., Jiao, Y., Carlisle, N. B., Xie, H., Fonzo, G. A., Zhang, Y. 2025; 184: 107066

  Abstract

  Neuroimaging techniques including functional magnetic resonance imaging (fMRI) and electroencephalogram (EEG) have shown promise in detecting functional abnormalities in various brain disorders. However, existing studies often focus on a single domain or modality, neglecting the valuable complementary information offered by multiple domains from both fMRI and EEG, which is crucial for a comprehensive representation of disorder pathology. This limitation poses a challenge in effectively leveraging the synergistic information derived from these modalities. To address this, we propose a Multi-modal Cross-domain Self-supervised Pre-training Model (MCSP), a novel approach that leverages self-supervised learning to synergize multi-modal information across spatial, temporal, and spectral domains. Our model employs cross-domain self-supervised loss that bridges domain differences by implementing domain-specific data augmentation and contrastive loss, enhancing feature discrimination. Furthermore, MCSP introduces cross-modal self-supervised loss to capitalize on the complementary information of fMRI and EEG, facilitating knowledge distillation within domains and maximizing cross-modal feature convergence. We constructed a large-scale pre-training dataset and pretrained MCSP model by leveraging proposed self-supervised paradigms to fully harness multimodal neuroimaging data. Through comprehensive experiments, we have demonstrated the superior performance and generalizability of our model on multiple classification tasks. Our study contributes a significant advancement in the fusion of fMRI and EEG, marking a novel integration of cross-domain features, which enriches the existing landscape of neuroimaging research, particularly within the context of mental disorder studies.

  View details for DOI 10.1016/j.neunet.2024.107066

  View details for PubMedID 39733703

  View details for PubMedCentralID PMC11802293

• Deep graph learning of multimodal brain networks defines treatment-predictive signatures in major depression. Molecular psychiatry Jiao, Y., Zhao, K., Wei, X., Carlisle, N. B., Keller, C. J., Oathes, D. J., Fonzo, G. A., Zhang, Y. 2025

  Abstract

  Major depressive disorder (MDD) presents a substantial health burden with low treatment response rates. Predicting antidepressant efficacy is challenging due to MDD's complex and varied neuropathology. Identifying biomarkers for antidepressant treatment requires thorough analysis of clinical trial data. Multimodal neuroimaging, combined with advanced data-driven methods, can enhance our understanding of the neurobiological processes influencing treatment outcomes. To address this, we analyzed resting-state fMRI and EEG connectivity data from 130 patients treated with sertraline and 135 patients with placebo from the Establishing Moderators and Biosignatures of Antidepressant Response in Clinical Care (EMBARC) study. A deep learning framework was developed using graph neural networks to integrate data-augmented connectivity and cross-modality correlation, aiming to predict individual symptom changes by revealing multimodal brain network signatures. The results showed that our model demonstrated promising prediction accuracy, with an R2 value of 0.24 for sertraline and 0.20 for placebo. It also exhibited potential in transferring predictions using only EEG. Key brain regions identified for predicting sertraline response included the inferior temporal gyrus (fMRI) and posterior cingulate cortex (EEG), while for placebo response, the precuneus (fMRI) and supplementary motor area (EEG) were critical. Additionally, both modalities identified the superior temporal gyrus and posterior cingulate cortex as significant for sertraline response, while the anterior cingulate cortex and postcentral gyrus were common predictors in the placebo arm. Additionally, variations in the frontoparietal control, ventral attention, dorsal attention, and limbic networks were notably associated with MDD treatment. By integrating fMRI and EEG, our study established novel multimodal brain network signatures to predict individual responses to sertraline and placebo in MDD, providing interpretable neural circuit patterns that may guide future targeted interventions. Trial Registration: Establishing Moderators and Biosignatures of Antidepressant Response for Clinical Care for Depression (EMBARC) ClinicalTrials.gov Identifier: NCT#01407094.

  View details for DOI 10.1038/s41380-025-02974-6

  View details for PubMedID 40164695

  View details for PubMedCentralID 6100771

• Diffusion transformer-augmented fMRI functional connectivity for enhanced autism spectrum disorder diagnosis. Journal of neural engineering Zhao, H., Lou, H., Yao, L., Zhang, Y. 2025; 22 (1)

  Abstract

  Objective.Functional magnetic resonance imaging (fMRI) is often modeled as networks of Regions of Interest and their functional connectivity to study brain functions and mental disorders. Limited fMRI data due to high acquisition costs hampers recognition model performance. We aim to address this issue using generative diffusion models for data augmentation.Approach.We propose Brain-Net-Diffusion, a transformer-based latent diffusion model to generate realistic functional connectivity for augmenting fMRI datasets and evaluate its impact on classification tasks.Main results.The Brain-Net-Diffusion effectively generates connectivity patterns resembling real data and significantly enhances classification performance. Augmentation using Brain-Net-Diffusion increased downstream autism spectrum disorder classification accuracy by 4.3% compared to no augmentation. It also outperformed other augmentation methods, with accuracy improvements ranging from 1.3% to 2.2%.Significance.Our approach demonstrates the effectiveness of diffusion models for fMRI data augmentation, providing a robust solution for overcoming data scarcity in functional connectivity analysis. To facilitate further research, we have made our code publicly available athttps://github.com/JoeZhao527/brain-net-diffusion.

  View details for DOI 10.1088/1741-2552/adb07a

  View details for PubMedID 39883961

• Multi-Modal Diagnosis of Alzheimer's Disease Using Interpretable Graph Convolutional Networks. IEEE transactions on medical imaging Zhou, H., He, L., Chen, B. Y., Shen, L., Zhang, Y. 2025; 44 (1): 142-153

  Abstract

  The interconnection between brain regions in neurological disease encodes vital information for the advancement of biomarkers and diagnostics. Although graph convolutional networks are widely applied for discovering brain connection patterns that point to disease conditions, the potential of connection patterns that arise from multiple imaging modalities has yet to be fully realized. In this paper, we propose a multi-modal sparse interpretable GCN framework (SGCN) for the detection of Alzheimer's disease (AD) and its prodromal stage, known as mild cognitive impairment (MCI). In our experimentation, SGCN learned the sparse regional importance probability to find signature regions of interest (ROIs), and the connective importance probability to reveal disease-specific brain network connections. We evaluated SGCN on the Alzheimer's Disease Neuroimaging Initiative database with multi-modal brain images and demonstrated that the ROI features learned by SGCN were effective for enhancing AD status identification. The identified abnormalities were significantly correlated with AD-related clinical symptoms. We further interpreted the identified brain dysfunctions at the level of large-scale neural systems and sex-related connectivity abnormalities in AD/MCI. The salient ROIs and the prominent brain connectivity abnormalities interpreted by SGCN are considerably important for developing novel biomarkers. These findings contribute to a better understanding of the network-based disorder via multi-modal diagnosis and offer the potential for precision diagnostics. The source code is available at https://github.com/Houliang-Zhou/SGCN.

  View details for DOI 10.1109/TMI.2024.3432531

  View details for PubMedID 39042528

  View details for PubMedCentralID PMC11754532

• CRISPR-GEM: A Novel Machine Learning Model for CRISPR Genetic Target Discovery and Evaluation. ACS synthetic biology Graham, J. P., Zhang, Y., He, L., Gonzalez-Fernandez, T. 2024; 13 (10): 3413-3429

  Abstract

  CRISPR gene editing strategies are shaping cell therapies through precise and tunable control over gene expression. However, limitations in safely delivering high quantities of CRISPR machinery demand careful target gene selection to achieve reliable therapeutic effects. Informed target gene selection requires a thorough understanding of the involvement of target genes in gene regulatory networks (GRNs) and thus their impact on cell phenotype. Effective decoding of these complex networks has been achieved using machine learning models, but current techniques are limited to single cell types and focus mainly on transcription factors, limiting their applicability to CRISPR strategies. To address this, we present CRISPR-GEM, a multilayer perceptron (MLP) based synthetic GRN constructed to accurately predict the downstream effects of CRISPR gene editing. First, input and output nodes are identified as differentially expressed genes between defined experimental and target cell/tissue types, respectively. Then, MLP training learns regulatory relationships in a black-box approach allowing accurate prediction of output gene expression using only input gene expression. Finally, CRISPR-mimetic perturbations are made to each input gene individually, and the resulting model predictions are compared to those for the target group to score and assess each input gene as a CRISPR candidate. The top scoring genes provided by CRISPR-GEM therefore best modulate experimental group GRNs to motivate transcriptomic shifts toward a target group phenotype. This machine learning model is the first of its kind for predicting optimal CRISPR target genes and serves as a powerful tool for enhanced CRISPR strategies across a range of cell therapies.

  View details for DOI 10.1021/acssynbio.4c00473

  View details for PubMedID 39375864

  View details for PubMedCentralID PMC11494708

• Evaluating the Quality of Brain MRI Generators. Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention Wu, J., Peng, W., Li, B., Zhang, Y., Pohl, K. M. 2024; 15010: 297-307

  Abstract

  Deep learning models generating structural brain MRIs have the potential to significantly accelerate discovery of neuroscience studies. However, their use has been limited in part by the way their quality is evaluated. Most evaluations of generative models focus on metrics originally designed for natural images (such as structural similarity index and Fréchet inception distance). As we show in a comparison of 6 state-of-the-art generative models trained and tested on over 3000 MRIs, these metrics are sensitive to the experimental setup and inadequately assess how well brain MRIs capture macrostructural properties of brain regions (a.k.a., anatomical plausibility). This shortcoming of the metrics results in inconclusive findings even when qualitative differences between the outputs of models are evident. We therefore propose a framework for evaluating models generating brain MRIs, which requires uniform processing of the real MRIs, standardizing the implementation of the models, and automatically segmenting the MRIs generated by the models. The segmentations are used for quantifying the plausibility of anatomy displayed in the MRIs. To ensure meaningful quantification, it is crucial that the segmentations are highly reliable. Our framework rigorously checks this reliability, a step often overlooked by prior work. Only 3 of the 6 generative models produced MRIs, of which at least 95% had highly reliable segmentations. More importantly, the assessment of each model by our framework is in line with qualitative assessments, reinforcing the validity of our approach. The code of this framework is available via https://github.com/jiaqiw01/MRIAnatEval.git.

  View details for DOI 10.1007/978-3-031-72117-5_28

  View details for PubMedID 40364898

  View details for PubMedCentralID PMC12066240

• XDL-ESI: Electrophysiological Sources Imaging via explainable deep learning framework with validation on simultaneous EEG and iEEG. NeuroImage Jiao, M., Xian, X., Wang, B., Zhang, Y., Yang, S., Chen, S., Sun, H., Liu, F. 2024; 299: 120802

  Abstract

  Electroencephalography (EEG) or Magnetoencephalography (MEG) source imaging aims to estimate the underlying activated brain sources to explain the observed EEG/MEG recordings. Solving the inverse problem of EEG/MEG Source Imaging (ESI) is challenging due to its ill-posed nature. To achieve a unique solution, it is essential to apply sophisticated regularization constraints to restrict the solution space. Traditionally, the design of regularization terms is based on assumptions about the spatiotemporal structure of the underlying source dynamics. In this paper, we propose a novel paradigm for ESI via an Explainable Deep Learning framework, termed as XDL-ESI, which connects the iterative optimization algorithm with deep learning architecture by unfolding the iterative updates with neural network modules. The proposed framework has the advantages of (1) establishing a data-driven approach to model the source solution structure instead of using hand-crafted regularization terms; (2) improving the robustness of source solutions by introducing a topological loss that leverages the geometric spatial information applying varying penalties on distinct localization errors; (3) improving the reconstruction efficiency and interpretability as it inherits the advantages from both the iterative optimization algorithms (interpretability) and deep learning approaches (function approximation). The proposed XDL-ESI framework provides an efficient, accurate, and interpretable paradigm to solve the ESI inverse problem with satisfactory performance in both simulated data and real clinical data. Specially, this approach is further validated using simultaneous EEG and intracranial EEG (iEEG).

  View details for DOI 10.1016/j.neuroimage.2024.120802

  View details for PubMedID 39173694

  View details for PubMedCentralID PMC11549933

• Dementia Subtypes Defined Through Neuropsychiatric Symptom-Associated Brain Connectivity Patterns. JAMA network open Zhao, K., Xie, H., Fonzo, G. A., Carlisle, N. B., Osorio, R. S., Zhang, Y. 2024; 7 (7): e2420479

  Abstract

  Understanding the heterogeneity of neuropsychiatric symptoms (NPSs) and associated brain abnormalities is essential for effective management and treatment of dementia.To identify dementia subtypes with distinct functional connectivity associated with neuropsychiatric subsyndromes.Using data from the Open Access Series of Imaging Studies-3 (OASIS-3; recruitment began in 2005) and Alzheimer Disease Neuroimaging Initiative (ADNI; recruitment began in 2004) databases, this cross-sectional study analyzed resting-state functional magnetic resonance imaging (fMRI) scans, clinical assessments, and neuropsychological measures of participants aged 42 to 95 years. The fMRI data were processed from July 2022 to February 2024, with secondary analysis conducted from August 2022 to March 2024. Participants without medical conditions or medical contraindications for MRI were recruited.A multivariate sparse canonical correlation analysis was conducted to identify functional connectivity-informed NPS subsyndromes, including behavioral and anxiety subsyndromes. Subsequently, a clustering analysis was performed on obtained latent connectivity profiles to reveal neurophysiological subtypes, and differences in abnormal connectivity and phenotypic profiles between subtypes were examined.Among 1098 participants in OASIS-3, 177 individuals who had fMRI and at least 1 NPS at baseline were included (78 female [44.1%]; median [IQR] age, 72 [67-78] years) as a discovery dataset. There were 2 neuropsychiatric subsyndromes identified: behavioral (r = 0.22; P = .002; P for permutation = .007) and anxiety (r = 0.19; P = .01; P for permutation = .006) subsyndromes from connectivity NPS-associated latent features. The behavioral subsyndrome was characterized by connections predominantly involving the default mode (within-network contribution by summed correlation coefficients = 54) and somatomotor (within-network contribution = 58) networks and NPSs involving nighttime behavior disturbance (R = -0.29; P < .001), agitation (R = -0.28; P = .001), and apathy (R = -0.23; P = .007). The anxiety subsyndrome mainly consisted of connections involving the visual network (within-network contribution = 53) and anxiety-related NPSs (R = 0.36; P < .001). By clustering individuals along these 2 subsyndrome-associated connectivity latent features, 3 subtypes were found (subtype 1: 45 participants; subtype 2: 43 participants; subtype 3: 66 participants). Patients with dementia of subtype 3 exhibited similar brain connectivity and cognitive behavior patterns to those of healthy individuals. However, patients with dementia of subtypes 1 and 2 had different dysfunctional connectivity profiles involving the frontoparietal control network (FPC) and somatomotor network (the difference by summed z values was 230 within the SMN and 173 between the SMN and FPC for subtype 1 and 473 between the SMN and visual network for subtype 2) compared with those of healthy individuals. These dysfunctional connectivity patterns were associated with differences in baseline dementia severity (eg, the median [IQR] of the total score of NPSs was 2 [2-7] for subtype 3 vs 6 [3-8] for subtype 1; P = .04 and 5.5 [3-11] for subtype 2; P = .03) and longitudinal progression of cognitive impairment and behavioral dysfunction (eg, the overall interaction association between time and subtypes to orientation was F = 4.88; P = .008; using the time × subtype 3 interaction item as the reference level: β = 0.05; t = 2.6 for time × subtype 2; P = .01). These findings were further validated using a replication dataset of 193 participants (127 female [65.8%]; median [IQR] age, 74 [69-77] years) consisting of 154 newly released participants from OASIS-3 and 39 participants from ADNI.These findings may provide a novel framework to disentangle the neuropsychiatric and brain functional heterogeneity of dementia, offering a promising avenue to improve clinical management and facilitate the timely development of targeted interventions for patients with dementia.

  View details for DOI 10.1001/jamanetworkopen.2024.20479

  View details for PubMedID 38976268

  View details for PubMedCentralID PMC11231801

• Adaptive Multimodel Knowledge Transfer Matrix Machine for EEG Classification. IEEE transactions on neural networks and learning systems Liang, S., Hang, W., Lei, B., Wang, J., Qin, J., Choi, K. S., Zhang, Y. 2024; 35 (6): 7726-7739

  Abstract

  The emerging matrix learning methods have achieved promising performances in electroencephalogram (EEG) classification by exploiting the structural information between the columns or rows of feature matrices. Due to the intersubject variability of EEG data, these methods generally need to collect a large amount of labeled individual EEG data, which would cause fatigue and inconvenience to the subjects. Insufficient subject-specific EEG data will weaken the generalization capability of the matrix learning methods in neural pattern decoding. To overcome this dilemma, we propose an adaptive multimodel knowledge transfer matrix machine (AMK-TMM), which can selectively leverage model knowledge from multiple source subjects and capture the structural information of the corresponding EEG feature matrices. Specifically, by incorporating least-squares (LS) loss with spectral elastic net regularization, we first present an LS support matrix machine (LS-SMM) to model the EEG feature matrices. To boost the generalization capability of LS-SMM in scenarios with limited EEG data, we then propose a multimodel adaption method, which can adaptively choose multiple correlated source model knowledge with a leave-one-out cross-validation strategy on the available target training data. We extensively evaluate our method on three independent EEG datasets. Experimental results demonstrate that our method achieves promising performances on EEG classification.

  View details for DOI 10.1109/TNNLS.2022.3220551

  View details for PubMedID 36383580

• Enabling temporal-spectral decoding in multi-class single-side upper limb classification ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE Jia, H., Han, S., Caiafa, C. F., Duan, F., Zhang, Y., Sun, Z., Sole-Casals, J. 2024; 133

  View details for DOI 10.1016/j.engappai.2024.108473

  View details for Web of Science ID 001237358700001

• Discriminative functional connectivity signature of cocaine use disorder links to rTMS treatment response. Nature. Mental health Zhao, K., Fonzo, G. A., Xie, H., Oathes, D. J., Keller, C. J., Carlisle, N. B., Etkin, A., Garza-Villarreal, E. A., Zhang, Y. 2024; 2 (4): 388-400

  Abstract

  Cocaine use disorder (CUD) is prevalent, and repetitive transcranial magnetic stimulation (rTMS) shows promise in reducing cravings. However, the association between a consistent CUD-specific functional connectivity signature and treatment response remains unclear. Here we identify a validated functional connectivity signature from functional magnetic resonance imaging to discriminate CUD, with successful independent replication. We found increased connectivity within the visual and dorsal attention networks and between the frontoparietal control and ventral attention networks, alongside reduced connectivity between the default mode and limbic networks in patients with CUD. These connections were associated with drug use history and cognitive impairments. Using data from a randomized clinical trial, we also established the prognostic value of these functional connectivities for rTMS treatment outcomes in CUD, especially involving the frontoparietal control and default mode networks. Our findings reveal insights into the neurobiological mechanisms of CUD and link functional connectivity biomarkers with rTMS treatment response, offering potential targets for future therapeutic development.

  View details for DOI 10.1038/s44220-024-00209-1

  View details for PubMedID 39279909

  View details for PubMedCentralID PMC11394333

• Symptom dimensions of resting-state electroencephalographic functional connectivity in autism. Nature. Mental health Tong, X., Xie, H., Fonzo, G. A., Zhao, K., Satterthwaite, T. D., Carlisle, N. B., Zhang, Y. 2024; 2 (3): 287-298

  Abstract

  Autism spectrum disorder (ASD) is a common neurodevelopmental disorder characterized by social and communication deficits (SCDs), restricted and repetitive behaviors (RRBs) and fixated interests. Despite its prevalence, development of effective therapy for ASD is hindered by its symptomatic and neurophysiological heterogeneities. To comprehensively explore these heterogeneities, we developed a new analytical framework combining contrastive learning and sparse canonical correlation analysis that identifies symptom-linked resting-state electroencephalographic connectivity dimensions within 392 ASD samples. We present two dimensions with multivariate connectivity basis exhibiting significant correlations with SCD and RRB, confirm their robustness through cross-validation and demonstrate their conceptual generalizability using an independent dataset (n = 222). Specifically, the right inferior parietal lobe is the core region for RRB, while connectivity between the left angular gyrus and the right middle temporal gyrus show key contribution to SCD. These findings provide a promising avenue to parse ASD heterogeneity with high clinical translatability, paving the way for ASD treatment development and precision medicine.

  View details for DOI 10.1038/s44220-023-00195-w

  View details for PubMedID 39219688

  View details for PubMedCentralID PMC11361313

• Early prediction of dementia using fMRI data with a graph convolutional network approach. Journal of neural engineering Han, S., Sun, Z., Zhao, K., Duan, F., Caiafa, C. F., Zhang, Y., Solé-Casals, J. 2024; 21 (1)

  Abstract

  Objective. Alzheimer's disease is a progressive neurodegenerative dementia that poses a significant global health threat. It is imperative and essential to detect patients in the mild cognitive impairment (MCI) stage or even earlier, enabling effective interventions to prevent further deterioration of dementia. This study focuses on the early prediction of dementia utilizing Magnetic Resonance Imaging (MRI) data, using the proposed Graph Convolutional Networks (GCNs).Approach. Specifically, we developed a functional connectivity (FC) based GCN framework for binary classifications using resting-state fMRI data. We explored different types and processing methods of FC and evaluated the performance on the OASIS-3 dataset. We developed the GCN model for two different purposes: (1) MCI diagnosis: classifying MCI from normal controls (NCs); and (2) dementia risk prediction: classifying NCs from subjects who have the potential for developing MCI but have not been clinically diagnosed as MCI.Main results. The results of the experiments revealed several important findings: First, the proposed GCN outperformed both the baseline GCN and Support Vector Machine (SVM). It achieved the best average accuracy of 80.3% (11.7% higher than the baseline GCN and 23.5% higher than SVM) and the highest accuracy of 91.2%. Secondly, the GCN framework with (absolute) individual FC performed slightly better than that with global FC generally. However, GCN using global graphs with appropriate connectivity can achieve equivalent or superior performance to individual graphs in some cases, which highlights the significance of suitable connectivity for achieving performance. Additionally, the results indicate that the self-network connectivity of specific brain network regions (such as default mode network, visual network, ventral attention network and somatomotor network) may play a more significant role in GCN classification.Significance. Overall, this study offers valuable insights into the application of GCNs in brain analysis and early diagnosis of dementia. This contributes significantly to the understanding of MCI and has substantial potential for clinical applications in early diagnosis and intervention for dementia and other neurodegenerative diseases. Our code for GCN implementation is available at:https://github.com/Shuning-Han/FC-based-GCN.

  View details for DOI 10.1088/1741-2552/ad1e22

  View details for PubMedID 38215493

• Sparse Bayesian Learning for End-to-End EEG Decoding. IEEE transactions on pattern analysis and machine intelligence Wang, W., Qi, F., Wipf, D. P., Cai, C., Yu, T., Li, Y., Zhang, Y., Yu, Z., Wu, W. 2023; 45 (12): 15632-15649

  Abstract

  Decoding brain activity from non-invasive electroencephalography (EEG) is crucial for brain-computer interfaces (BCIs) and the study of brain disorders. Notably, end-to-end EEG decoding has gained widespread popularity in recent years owing to the remarkable advances in deep learning research. However, many EEG studies suffer from limited sample sizes, making it difficult for existing deep learning models to effectively generalize to highly noisy EEG data. To address this fundamental limitation, this paper proposes a novel end-to-end EEG decoding algorithm that utilizes a low-rank weight matrix to encode both spatio-temporal filters and the classifier, all optimized under a principled sparse Bayesian learning (SBL) framework. Importantly, this SBL framework also enables us to learn hyperparameters that optimally penalize the model in a Bayesian fashion. The proposed decoding algorithm is systematically benchmarked on five motor imagery BCI EEG datasets ( N=192) and an emotion recognition EEG dataset ( N=45), in comparison with several contemporary algorithms, including end-to-end deep-learning-based EEG decoding algorithms. The classification results demonstrate that our algorithm significantly outperforms the competing algorithms while yielding neurophysiologically meaningful spatio-temporal patterns. Our algorithm therefore advances the state-of-the-art by providing a novel EEG-tailored machine learning tool for decoding brain activity.

  View details for DOI 10.1109/TPAMI.2023.3299568

  View details for PubMedID 37506000

• Assessing the Potential of Data Augmentation in EEG Functional Connectivity for Early Detection of Alzheimer's Disease COGNITIVE COMPUTATION Jia, H., Huang, Z., Caiafa, C. F., Duan, F., Zhang, Y., Sun, Z., Sole-Casals, J. 2024; 16 (1): 229-242

  View details for DOI 10.1007/s12559-023-10188-7

  View details for Web of Science ID 001070115900001

• Multi-Class Classification of Upper Limb Movements With Filter Bank Task-Related Component Analysis. IEEE journal of biomedical and health informatics Jia, H., Feng, F., Caiafa, C. F., Duan, F., Zhang, Y., Sun, Z., Sole-Casals, J. 2023; 27 (8): 3867-3877

  Abstract

  The classification of limb movements can provide with control commands in non-invasive brain-computer interface. Previous studies on the classification of limb movements have focused on the classification of left/right limbs; however, the classification of different types of upper limb movements has often been ignored despite that it provides more active-evoked control commands in the brain-computer interface. Nevertheless, few machine learning method can be used as the state-of-the-art method in the multi-class classification of limb movements. This work focuses on the multi-class classification of upper limb movements and proposes the multi-class filter bank task-related component analysis (mFBTRCA) method, which consists of three steps: spatial filtering, similarity measuring and filter bank selection. The spatial filter, namely the task-related component analysis, is first used to remove noise from EEG signals. The canonical correlation measures the similarity of the spatial-filtered signals and is used for feature extraction. The correlation features are extracted from multiple low-frequency filter banks. The minimum-redundancy maximum-relevance selects the essential features from all the correlation features, and finally, the support vector machine is used to classify the selected features. The proposed method compared against previously used models is evaluated using two datasets. mFBTRCA achieved a classification accuracy of 0.4193 ± 0.0780 (7 classes) and 0.4032 ± 0.0714 (5 classes), respectively, which improves on the best accuracies achieved using the compared methods (0.3590 ± 0.0645 and 0.3159 ± 0.0736, respectively). The proposed method is expected to provide more control commands in the applications of non-invasive brain-computer interfaces.

  View details for DOI 10.1109/JBHI.2023.3278747

  View details for PubMedID 37227915

• Structure invariance-driven collaborative contrastive network for EEG decoding BIOMEDICAL SIGNAL PROCESSING AND CONTROL Liang, S., Li, Z., Hang, W., Zhang, Y., Lei, B., Qin, J., Zhang, Y., Choi, K. 2023; 86

  View details for DOI 10.1016/j.bspc.2023.105214

  View details for Web of Science ID 001031696000001

• Recent Approaches to Design and Analysis of Electrical Impedance Systems for Single Cells Using Machine Learning. Sensors (Basel, Switzerland) Ferguson, C., Zhang, Y., Palego, C., Cheng, X. 2023; 23 (13)

  Abstract

  Individual cells have many unique properties that can be quantified to develop a holistic understanding of a population. This can include understanding population characteristics, identifying subpopulations, or elucidating outlier characteristics that may be indicators of disease. Electrical impedance measurements are rapid and label-free for the monitoring of single cells and generate large datasets of many cells at single or multiple frequencies. To increase the accuracy and sensitivity of measurements and define the relationships between impedance and biological features, many electrical measurement systems have incorporated machine learning (ML) paradigms for control and analysis. Considering the difficulty capturing complex relationships using traditional modelling and statistical methods due to population heterogeneity, ML offers an exciting approach to the systemic collection and analysis of electrical properties in a data-driven way. In this work, we discuss incorporation of ML to improve the field of electrical single cell analysis by addressing the design challenges to manipulate single cells and sophisticated analysis of electrical properties that distinguish cellular changes. Looking forward, we emphasize the opportunity to build on integrated systems to address common challenges in data quality and generalizability to save time and resources at every step in electrical measurement of single cells.

  View details for DOI 10.3390/s23135990

  View details for PubMedID 37447838

  View details for PubMedCentralID PMC10346645

• Individualized fMRI connectivity defines signatures of antidepressant and placebo responses in major depression. Molecular psychiatry Zhao, K., Xie, H., Fonzo, G. A., Tong, X., Carlisle, N., Chidharom, M., Etkin, A., Zhang, Y. 2023; 28 (6): 2490-2499

  Abstract

  Though sertraline is commonly prescribed in patients with major depressive disorder (MDD), its superiority over placebo is only marginal. This is in part due to the neurobiological heterogeneity of the individuals. Characterizing individual-unique functional architecture of the brain may help better dissect the heterogeneity, thereby defining treatment-predictive signatures to guide personalized medication. In this study, we investigate whether individualized brain functional connectivity (FC) can define more predictable signatures of antidepressant and placebo treatment in MDD. The data used in the present work were collected by the Establishing Moderators and Biosignatures of Antidepressant Response in Clinical Care (EMBARC) study. Patients (N = 296) were randomly assigned to antidepressant sertraline or placebo double-blind treatment for 8 weeks. The whole-brain FC networks were constructed from pre-treatment resting-state functional magnetic resonance imaging (rs-fMRI). Then, FC was individualized by removing the common components extracted from the raw baseline FC to train regression-based connectivity predictive models. With individualized FC features, the established prediction models successfully identified signatures that explained 22% variance for the sertraline group and 31% variance for the placebo group in predicting HAMD17 change. Compared with the raw FC-based models, the individualized FC-defined signatures significantly improved the prediction performance, as confirmed by cross-validation. For sertraline treatment, predictive FC metrics were predominantly located in the left middle temporal cortex and right insula. For placebo, predictive FC metrics were primarily located in the bilateral cingulate cortex and left superior temporal cortex. Our findings demonstrated that through the removal of common FC components, individualization of FC metrics enhanced the prediction performance compared to raw FC. Associated with previous MDD clinical studies, our identified predictive biomarkers provided new insights into the neuropathology of antidepressant and placebo treatment.

  View details for DOI 10.1038/s41380-023-01958-8

  View details for PubMedID 36732585

  View details for PubMedCentralID 5623437

• Individual Deviations from Normative Electroencephalographic Connectivity Predict Antidepressant Response. medRxiv : the preprint server for health sciences Tong, X., Xie, H., Wu, W., Keller, C., Fonzo, G., Chidharom, M., Carlisle, N., Etkin, A., Zhang, Y. 2023

  Abstract

  Antidepressant medications yield unsatisfactory treatment outcomes in patients with major depressive disorder (MDD) with modest advantages over the placebo. This modest efficacy is partly due to the elusive mechanisms of antidepressant responses and unexplained heterogeneity in patient's response to treatment - the approved antidepressants only benefit a portion of patients, calling for personalized psychiatry based on individual-level prediction of treatment responses. Normative modeling, a framework that quantifies individual deviations in psychopathological dimensions, offers a promising avenue for the personalized treatment for psychiatric disorders. In this study, we built a normative model with resting-state electroencephalography (EEG) connectivity data from healthy controls of three independent cohorts. We characterized the individual deviation of MDD patients from the healthy norms, based on which we trained sparse predictive models for treatment responses of MDD patients. We successfully predicted treatment outcomes for patients receiving sertraline (r = 0.43, p < 0.001) and placebo (r = 0.33, p < 0.001). We also showed that the normative modeling framework successfully distinguished subclinical and diagnostic variabilities among subjects. From the predictive models, we identified key connectivity signatures in resting-state EEG for antidepressant treatment, suggesting differences in neural circuit involvement between treatment responses. Our findings and highly generalizable framework advance the neurobiological understanding in the potential pathways of antidepressant responses, enabling more targeted and effective MDD treatment.

  View details for DOI 10.1101/2023.05.24.23290434

  View details for PubMedID 37292874

  View details for PubMedCentralID PMC10246152

• A generalizable functional connectivity signature characterizes brain dysfunction and links to rTMS treatment response in cocaine use disorder. medRxiv : the preprint server for health sciences Zhao, K., Fonzo, G. A., Xie, H., Oathes, D. J., Keller, C. J., Carlisle, N., Etkin, A., Garza-Villarreal, E. A., Zhang, Y. 2023

  Abstract

  Cocaine use disorder (CUD) is a prevalent substance abuse disorder, and repetitive transcranial magnetic stimulation (rTMS) has shown potential in reducing cocaine cravings. However, a robust and replicable biomarker for CUD phenotyping is lacking, and the association between CUD brain phenotypes and treatment response remains unclear. Our study successfully established a cross-validated functional connectivity signature for accurate CUD phenotyping, using resting-state functional magnetic resonance imaging from a discovery cohort, and demonstrated its generalizability in an independent replication cohort. We identified phenotyping FCs involving increased connectivity between the visual network and dorsal attention network, and between the frontoparietal control network and ventral attention network, as well as decreased connectivity between the default mode network and limbic network in CUD patients compared to healthy controls. These abnormal connections correlated significantly with other drug use history and cognitive dysfunctions, e.g., non-planning impulsivity. We further confirmed the prognostic potential of the identified discriminative FCs for rTMS treatment response in CUD patients and found that the treatment-predictive FCs mainly involved the frontoparietal control and default mode networks. Our findings provide new insights into the neurobiological mechanisms of CUD and the association between CUD phenotypes and rTMS treatment response, offering promising targets for future therapeutic development.

  View details for DOI 10.1101/2023.04.21.23288948

  View details for PubMedID 37162878

  View details for PubMedCentralID PMC10168499

• Less Is More: Brain Functional Connectivity Empowered Generalizable Intention Classification With Task-Relevant Channel Selection. IEEE transactions on neural systems and rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society Lou, H., Ye, Z., Yao, L., Zhang, Y. 2023; 31: 1888-1899

  Abstract

  Electroencephalography (EEG) signals are gaining popularity in Brain-Computer Interface (BCI)-based rehabilitation and neural engineering applications thanks to their portability and availability. Inevitably, the sensory electrodes on the entire scalp would collect signals irrelevant to the particular BCI task, increasing the risks of overfitting in machine learning-based predictions. While this issue is being addressed by scaling up the EEG datasets and handcrafting the complex predictive models, this also leads to increased computation costs. Moreover, the model trained for one set of subjects cannot easily be adapted to other sets due to inter-subject variability, which creates even higher over-fitting risks. Meanwhile, despite previous studies using either convolutional neural networks (CNNs) or graph neural networks (GNNs) to determine spatial correlations between brain regions, they fail to capture brain functional connectivity beyond physical proximity. To this end, we propose 1) removing task-irrelevant noises instead of merely complicating models; 2) extracting subject-invariant discriminative EEG encodings, by taking functional connectivity into account. Specifically, we construct a task-adaptive graph representation of the brain network based on topological functional connectivity rather than distance-based connections. Further, non-contributory EEG channels are excluded by selecting only functional regions relevant to the corresponding intention. We empirically show that the proposed approach outperforms the state-of-the-art, with around 1% and 11% improvements over CNN-based and GNN-based models, on performing motor imagery predictions. Also, the task-adaptive channel selection demonstrates similar predictive performance with only 20% of raw EEG data, suggesting a possible shift in direction for future works other than simply scaling up the model.

  View details for DOI 10.1109/TNSRE.2023.3252610

  View details for PubMedID 37028028

• Privacy-Preserving Multi-Source Domain Adaptation for Medical Data. IEEE journal of biomedical and health informatics Han, T., Gong, X., Feng, F., Zhang, J., Sun, Z., Zhang, Y. 2023; 27 (2): 842-853

  Abstract

  Great progress has been made in diagnosing medical diseases based on deep learning. Large-scale medical data are expected to improve deep learning performance further. It is almost impossible for a single institution to collect so much data due to the time-consuming and costly collection and labeling of medical data. Many studies have turned attention to data sharing among multiple medical institutions. However, due to different data acquiring and processing procedures, multiple institutions' medical data is characterized by distribution heterogeneity. Besides, the protection of patient privacy in medical data sharing has also been a common concern. To simultaneously address the problems of heterogeneous data distribution and privacy protection, we propose a novel multi-source source free domain adaptation. When aligning distributed heterogeneous data, our method only require to transfer the pre-trained source models rather than the direct source domain data, thus protecting patients' privacy. In addition, it has the advantages of being efficient and less costly in network resources. The proposed method is evaluated on the multi-site fMRI database Autism Brain Imaging Data Exchange (ABIDE) and yields an average accuracy of 69.37%. We also analyzed its effectiveness on network resource-saving and conducted additional experiments on Camelyon17 to validate the generalization.

  View details for DOI 10.1109/JBHI.2022.3175071

  View details for PubMedID 35594226

• Editorial: Deep learning in neuroimaging-based neurological disease analysis. Frontiers in neuroimaging Li, X., Zhang, Y., Zhao, Q. 2023; 2: 1127719

  View details for DOI 10.3389/fnimg.2023.1127719

  View details for PubMedID 37554630

  View details for PubMedCentralID PMC10406280

• Single-Cell Classification Based on Population Nucleus Size Combining Microwave Impedance Spectroscopy and Machine Learning. Sensors (Basel, Switzerland) Ferguson, C. A., Hwang, J. C., Zhang, Y., Cheng, X. 2023; 23 (2)

  Abstract

  Many recent efforts in the diagnostic field address the accessibility of cancer diagnosis. Typical histological staining methods identify cancer cells visually by a larger nucleus with more condensed chromatin. Machine learning (ML) has been incorporated into image analysis for improving this process. Recently, impedance spectrometers have been shown to generate all-inclusive lab-on-a-chip platforms to detect nucleus abnormities. In this paper, a wideband electrical sensor and data analysis paradigm that can identify nuclear changes shows the realization of a single-cell microfluidic device to detect nuclei of altered sizes. To model cells of altered nucleus, Jurkat cells were treated to enlarge or shrink their nucleus followed by broadband sensing to obtain the S-parameters of single cells. The ability to deduce important frequencies associated with nucleus size is demonstrated and used to improve classification models in both binary and multiclass scenarios, despite a heterogeneous and overlapping cell population. The important frequency features match those predicted in a double-shell circuit model published in prior work, demonstrating a coherent new analytical technique for electrical data analysis. The electrical sensing platform assisted by ML with impressive accuracy of cell classification looks forward to a label-free and flexible approach to cancer diagnosis.

  View details for DOI 10.3390/s23021001

  View details for PubMedID 36679798

  View details for PubMedCentralID PMC9860723

• Machine learning-based identification of a psychotherapy-predictive electroencephalographic signature in PTSD Nature Mental Health Zhang, Y., Naparstek, S., Gordon, J., Watts, M., Shpigel, E., El-Said, D., Badami, F. S., Eisenberg, M. L., Toll, R. T., Gage, A., Goodkind, M. S., Etkin, A., Wu, W. 2023; 1: 284-294

  View details for DOI 10.1038/s44220-023-00049-5

• NORMATIVE MODELING VIA CONDITIONAL VARIATIONAL AUTOENCODER AND ADVERSARIAL LEARNING TO IDENTIFY BRAIN DYSFUNCTION IN ALZHEIMER'S DISEASE Wang, X., Zhou, R., Zhao, K., Leow, A., Zhang, Y., He, L., IEEE IEEE. 2023

  View details for DOI 10.1109/ISBI53787.2023.10230377

  View details for Web of Science ID 001062050500055

• Electrophysiological Brain Source Imaging via Combinatorial Search with Provable Optimality Wan, G., Jiao, M., Ju, X., Zhang, Y., Schweitzer, H., Liu, F., Williams, B., Chen, Y., Neville, J. ASSOC ADVANCEMENT ARTIFICIAL INTELLIGENCE. 2023: 12491-12499

  View details for Web of Science ID 001243749200111

• Attentive Deep Canonical Correlation Analysis for Diagnosing Alzheimer's Disease Using Multimodal Imaging Genetics Zhou, R., Zhou, H., Chen, B. Y., Shen, L., Zhang, Y., He, L., Greenspan, H., Madabhushi, A., Mousavi, P., Salcudean, S., Duncan, J., Syeda-Mahmood, T., Taylor, R. SPRINGER INTERNATIONAL PUBLISHING AG. 2023: 681-691

  View details for DOI 10.1007/978-3-031-43895-0_64

  View details for Web of Science ID 001109624900064

• Cerebral cortex layer segmentation using diffusion magnetic resonance imaging in vivo with applications to laminar connections and working memory analysis. Human brain mapping Zhang, J., Sun, Z., Duan, F., Shi, L., Zhang, Y., Solé-Casals, J., Caiafa, C. F. 2022; 43 (17): 5220-5234

  Abstract

  Understanding the laminar brain structure is of great help in further developing our knowledge of the functions of the brain. However, since most layer segmentation methods are invasive, it is difficult to apply them to the human brain in vivo. To systematically explore the human brain's laminar structure noninvasively, the K-means clustering algorithm was used to automatically segment the left hemisphere into two layers, the superficial and deep layers, using a 7 Tesla (T) diffusion magnetic resonance imaging (dMRI)open dataset. The obtained layer thickness was then compared with the layer thickness of the BigBrain reference dataset, which segmented the neocortex into six layers based on the von Economo atlas. The results show a significant correlation not only between our automatically segmented superficial layer thickness and the thickness of layers 1-3 from the reference histological data, but also between our automatically segmented deep layer thickness and the thickness of layers 4-6 from the reference histological data. Second, we constructed the laminar connections between two pairs of unidirectional connected regions, which is consistent with prior research. Finally, we conducted the laminar analysis of the working memory, which was challenging to do in the past, and explained the conclusions of the functional analysis. Our work successfully demonstrates that it is possible to segment the human cortex noninvasively into layers using dMRI data and further explores the mechanisms of the human brain.

  View details for DOI 10.1002/hbm.25998

  View details for PubMedID 35778791

  View details for PubMedCentralID PMC9812233

• A Novel Method for Constructing EEG Large-Scale Cortical Dynamical Functional Network Connectivity (dFNC): WTCS. IEEE transactions on cybernetics Yi, C., Yao, R., Song, L., Jiang, L., Si, Y., Li, P., Li, F., Yao, D., Zhang, Y., Xu, P. 2022; 52 (12): 12869-12881

  Abstract

  As a kind of biological network, the brain network conduces to understanding the mystery of high-efficiency information processing in the brain, which will provide instructions to develop efficient brain-like neural networks. Large-scale dynamical functional network connectivity (dFNC) provides a more context-sensitive, dynamical, and straightforward sight at a higher network level. Nevertheless, dFNC analysis needs good enough resolution in both temporal and spatial domains, and the construction of dFNC needs to capture the time-varying correlations between two multivariate time series with unmatched spatial dimensions. Effective methods still lack. With well-developed source imaging techniques, electroencephalogram (EEG) has the potential to possess both high temporal and spatial resolutions. Therefore, we proposed to construct the EEG large-scale cortical dFNC based on brain atlas to probe the subtle dynamic activities in the brain and developed a novel method, that is, wavelet coherence-S estimator (WTCS), to assess the dynamic couplings among functional subnetworks with different spatial dimensions. The simulation study demonstrated its robustness and availability of applying to dFNC. The application in real EEG data revealed the appealing "Primary peak" and "P3-like peak" in dFNC network properties and meaningful evolutions in dFNC network topology for P300. Our study brings new insights for probing brain activities at a more dynamical and higher hierarchical level and pushing forward the development of brain-inspired artificial neural networks. The proposed WTCS not only benefits the dFNC studies but also gives a new solution to capture the time-varying couplings between the multivariate time series that is often encountered in signal processing disciplines.

  View details for DOI 10.1109/TCYB.2021.3090770

  View details for PubMedID 34398778

• Improving pre-movement pattern detection with filter bank selection. Journal of neural engineering Jia, H., Sun, Z., Duan, F., Zhang, Y., Caiafa, C. F., Solé-Casals, J. 2022; 19 (6)

  Abstract

  Objective. Pre-movement decoding plays an important role in detecting the onsets of actions using low-frequency electroencephalography (EEG) signals before the movement of an upper limb. In this work, a binary classification method is proposed between two different states.Approach. The proposed method, referred to as filter bank standard task-related component analysis (FBTRCA), is to incorporate filter bank selection into the standard task-related component analysis (STRCA) method. In FBTRCA, the EEG signals are first divided into multiple sub-bands which start at specific fixed frequencies and end frequencies that follow in an arithmetic sequence. The STRCA method is then applied to the EEG signals in these bands to extract CCPs. The minimum redundancy maximum relevance feature selection method is used to select essential features from these correlation patterns in all sub-bands. Finally, the selected features are classified using the binary support vector machine classifier. A convolutional neural network (CNN) is an alternative approach to select canonical correlation patterns.Main Results. Three methods were evaluated using EEG signals in the time window from 2 s before the movement onset to 1 s after the movement onset. In the binary classification between a movement state and the resting state, the FBTRCA achieved an average accuracy of 0.8968 ± 0.0847 while the accuracies of STRCA and CNN were 0.8228 ± 0.1149 and 0.8828 ± 0.0917, respectively. In the binary classification between two actions, the accuracies of STRCA, CNN, and FBTRCA were 0.6611 ± 0.1432, 0.6993 ± 0.1271, 0.7178 ± 0.1274, respectively. Feature selection using filter banks, as in FBTRCA, produces comparable results to STRCA.Significance. The proposed method provides a way to select filter banks in pre-movement decoding, and thus it improves the classification performance. The improved pre-movement decoding of single upper limb movements is expected to provide people with severe motor disabilities with a more natural, non-invasive control of their external devices.

  View details for DOI 10.1088/1741-2552/ac9e75

  View details for PubMedID 36317288

• Modern views of machine learning for precision psychiatry. Patterns (New York, N.Y.) Chen, Z. S., Kulkarni, P. P., Galatzer-Levy, I. R., Bigio, B., Nasca, C., Zhang, Y. 2022; 3 (11): 100602

  Abstract

  In light of the National Institute of Mental Health (NIMH)'s Research Domain Criteria (RDoC), the advent of functional neuroimaging, novel technologies and methods provide new opportunities to develop precise and personalized prognosis and diagnosis of mental disorders. Machine learning (ML) and artificial intelligence (AI) technologies are playing an increasingly critical role in the new era of precision psychiatry. Combining ML/AI with neuromodulation technologies can potentially provide explainable solutions in clinical practice and effective therapeutic treatment. Advanced wearable and mobile technologies also call for the new role of ML/AI for digital phenotyping in mobile mental health. In this review, we provide a comprehensive review of ML methodologies and applications by combining neuroimaging, neuromodulation, and advanced mobile technologies in psychiatry practice. We further review the role of ML in molecular phenotyping and cross-species biomarker identification in precision psychiatry. We also discuss explainable AI (XAI) and neuromodulation in a closed human-in-the-loop manner and highlight the ML potential in multi-media information extraction and multi-modal data fusion. Finally, we discuss conceptual and practical challenges in precision psychiatry and highlight ML opportunities in future research.

  View details for DOI 10.1016/j.patter.2022.100602

  View details for PubMedID 36419447

  View details for PubMedCentralID PMC9676543

• An efficient CNN-LSTM network with spectral normalization and label smoothing technologies for SSVEP frequency recognition. Journal of neural engineering Pan, Y., Chen, J., Zhang, Y., Zhang, Y. 2022; 19 (5)

  Abstract

  Objective. Steady-state visual evoked potentials (SSVEPs) based brain-computer interface (BCI) has received great interests owing to the high information transfer rate and available large number of targets. However, the performance of frequency recognition methods heavily depends on the amount of the calibration data for intra-subject classification. Some research adopted the deep learning (DL) algorithm to conduct the inter-subject classification, which could reduce the calculation procedure, but the performance still has large room to improve compared with the intra-subject classification.Approach. To address these issues, we proposed an efficient SSVEP DL NETwork (termed SSVEPNET) based on one-dimensional convolution and long short-term memory (LSTM) module. To enhance the performance of SSVEPNET, we adopted the spectral normalization and label smoothing technologies during implementing the network architecture. We evaluated the SSVEPNET and compared it with other methods for the intra- and inter-subject classification under different conditions, i.e. two datasets, two time-window lengths (1 s and 0.5 s), three sizes of training data.Main results. Under all the experimental settings, the proposed SSVEPNET achieved the highest average accuracy for the intra- and inter-subject classification on the two SSVEP datasets, when compared with other traditional and DL baseline methods.Significance. The extensive experimental results demonstrate that the proposed DL model holds promise to enhance frequency recognition performance in SSVEP-based BCIs. Besides, the mixed network structures with convolutional neural network and LSTM, and the spectral normalization and label smoothing could be useful optimization strategies to design efficient models for electroencephalography data.

  View details for DOI 10.1088/1741-2552/ac8dc5

  View details for PubMedID 36041426

• Transdiagnostic connectome signatures from resting-state fMRI predict individual-level intellectual capacity. Translational psychiatry Tong, X., Xie, H., Carlisle, N., Fonzo, G. A., Oathes, D. J., Jiang, J., Zhang, Y. 2022; 12 (1): 367

  Abstract

  Medication and other therapies for psychiatric disorders show unsatisfying efficacy, in part due to the significant clinical/ biological heterogeneity within each disorder and our over-reliance on categorical clinical diagnoses. Alternatively, dimensional transdiagnostic studies have provided a promising pathway toward realizing personalized medicine and improved treatment outcomes. One factor that may influence response to psychiatric treatments is cognitive function, which is reflected in one's intellectual capacity. Intellectual capacity is also reflected in the organization and structure of intrinsic brain networks. Using a large transdiagnostic cohort (n = 1721), we sought to discover neuroimaging biomarkers by developing a resting-state functional connectome-based prediction model for a key intellectual capacity measure, Full-Scale Intelligence Quotient (FSIQ), across the diagnostic spectrum. Our cross-validated model yielded an excellent prediction accuracy (r = 0.5573, p < 0.001). The robustness and generalizability of our model was further validated on three independent cohorts (n = 2641). We identified key transdiagnostic connectome signatures underlying FSIQ capacity involving the dorsal-attention, frontoparietal and default-mode networks. Meanwhile, diagnosis groups showed disorder-specific biomarker patterns. Our findings advance the neurobiological understanding of cognitive functioning across traditional diagnostic categories and provide a new avenue for neuropathological classification of psychiatric disorders.

  View details for DOI 10.1038/s41398-022-02134-2

  View details for PubMedID 36068228

  View details for PubMedCentralID PMC9448815

• Sparse Interpretation of Graph Convolutional Networks for Multi-Modal Diagnosis of Alzheimer's Disease. Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention Zhou, H., Zhang, Y., Chen, B. Y., Shen, L., He, L. 2022; 13438: 469-478

  Abstract

  The interconnected quality of brain regions in neurological disease has immense importance for the development of biomarkers and diagnostics. While Graph Convolutional Network (GCN) methods are fundamentally compatible with discovering the connected role of brain regions in disease, current methods apply limited consideration for node features and their connectivity in brain network analysis. In this paper, we propose a sparse interpretable GCN framework (SGCN) for the identification and classification of Alzheimer's disease (AD) using brain imaging data with multiple modalities. SGCN applies an attention mechanism with sparsity to identify the most discriminative subgraph structure and important node features for the detection of AD. The model learns the sparse importance probabilities for each node feature and edge with entropy, ℓ 1, and mutual information regularization. We then utilized this information to find signature regions of interest (ROIs), and emphasize the disease-specific brain network connections by detecting the significant difference of connectives between regions in healthy control (HC), and AD groups. We evaluated SGCN on the ADNI database with imaging data from three modalities, including VBM-MRI, FDG-PET, and AV45-PET, and observed that the important probabilities it learned are effective for disease status identification and the sparse interpretability of disease-specific ROI features and connections. The salient ROIs detected and the most discriminative network connections interpreted by our method show a high correspondence with previous neuroimaging evidence associated with AD.

  View details for DOI 10.1007/978-3-031-16452-1_45

  View details for PubMedID 36827208

  View details for PubMedCentralID PMC9942706

• Motor Imagery Decoding in the Presence of Distraction Using Graph Sequence Neural Networks. IEEE transactions on neural systems and rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society Cai, S., Li, H., Wu, Q., Liu, J., Zhang, Y. 2022; 30: 1716-1726

  Abstract

  In this study, we propose a graph sequence neural network (GSNN) to accurately decode patterns of motor imagery from electroencephalograms (EEGs) in the presence of distractions. GSNN aims to build subgraphs by exploiting biological topologies among brain regions to capture local and global relationships across characteristic channels. Specifically, we model the similarity between pairwise EEG channels by the adjacency matrix of the graph sequence neural network. In addition, we propose a node domain attention selection network in which the connection and sparsity of the adjacency matrix can be adjusted dynamically according to the EEG signals acquired from different subjects. Extensive experiments on the public Berlin-distraction dataset show that in most experimental settings, our model performs considerably better than the state-of-the-art models. Moreover, comparative experiments indicate that our proposed node domain attention selection network plays a crucial role in improving the sensibility and adaptability of the GSNN model. The results show that the GSNN algorithm obtained superior classification accuracy (The average value of Recall, Precision, and F-score were 80.44%, 81.07% and 80.54%) compared to the state-of-the-art models. Finally, in the process of extracting the intermediate results, the relationships between important brain regions and channels were revealed to different influences in distraction themes.

  View details for DOI 10.1109/TNSRE.2022.3183023

  View details for PubMedID 35700243

• A dynamic graph convolutional neural network framework reveals new insights into connectome dysfunctions in ADHD. NeuroImage Zhao, K., Duka, B., Xie, H., Oathes, D. J., Calhoun, V., Zhang, Y. 2022; 246: 118774

  Abstract

  The pathological mechanism of attention deficit hyperactivity disorder (ADHD) is incompletely specified, which leads to difficulty in precise diagnosis. Functional magnetic resonance imaging (fMRI) has emerged as a common neuroimaging technique for studying the brain functional connectome. Most existing methods that have either ignored or simply utilized graph structure, do not fully leverage the potentially important topological information which may be useful in characterizing brain disorders. There is a crucial need for designing novel and efficient approaches which can capture such information. To this end, we propose a new dynamic graph convolutional network (dGCN), which is trained with sparse brain regional connections from dynamically calculated graph features. We also develop a novel convolutional readout layer to improve graph representation. Our extensive experimental analysis demonstrates significantly improved performance of dGCN for ADHD diagnosis compared with existing machine learning and deep learning methods. Visualizations of the salient regions of interest (ROIs) and connectivity based on informative features learned by our model show that the identified functional abnormalities mainly involve brain regions in temporal pole, gyrus rectus, and cerebellar gyri from temporal lobe, frontal lobe, and cerebellum, respectively. A positive correlation was further observed between the identified connectomic abnormalities and ADHD symptom severity. The proposed dGCN model shows great promise in providing a functional network-based precision diagnosis of ADHD and is also broadly applicable to brain connectome-based study of mental disorders.

  View details for DOI 10.1016/j.neuroimage.2021.118774

  View details for PubMedID 34861391

  View details for PubMedCentralID PMC10569447

• Deep EEG feature learning via stacking common spatial pattern and support matrix machine BIOMEDICAL SIGNAL PROCESSING AND CONTROL Liang, S., Hang, W., Yin, M., Shen, H., Wang, Q., Qin, J., Choi, K., Zhang, Y. 2022; 74

  View details for DOI 10.1016/j.bspc.2022.103531

  View details for Web of Science ID 000782646000003

• INTERPRETABLE GRAPH CONVOLUTIONAL NETWORK OF MULTI-MODALITY BRAIN IMAGING FOR ALZHEIMER'S DISEASE DIAGNOSIS Zhou, H., He, L., Zhang, Y., Shen, L., Chen, B., IEEE IEEE. 2022

  View details for DOI 10.1109/ISBI52829.2022.9761449

  View details for Web of Science ID 000836243800050

• Contrastive Functional Connectivity Graph Learning for Population-based fMRI Classification Wang, X., Yao, L., Rekik, I., Zhang, Y., Wang, L., Dou, Q., Fletcher, P. T., Speidel, S., Li, S. SPRINGER INTERNATIONAL PUBLISHING AG. 2022: 221-230

  View details for DOI 10.1007/978-3-031-16431-6_21

  View details for Web of Science ID 000867524300021

• Side-Aware Meta-Learning for Cross-Dataset Listener Diagnosis With Subjective Tinnitus IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING Liu, Z., Li, Y., Yao, L., Lucas, M., Monaghan, J. M., Zhang, Y. 2022; 30: 2352-2361

  Abstract

  With the development of digital technology, machine learning has paved the way for the next generation of tinnitus diagnoses. Although machine learning has been widely applied in EEG-based tinnitus analysis, most current models are dataset-specific. Each dataset may be limited to a specific range of symptoms, overall disease severity, and demographic attributes; further, dataset formats may differ, impacting model performance. This paper proposes a side-aware meta-learning for cross-dataset tinnitus diagnosis, which can effectively classify tinnitus in subjects of divergent ages and genders from different data collection processes. Owing to the superiority of meta-learning, our method does not rely on large-scale datasets like conventional deep learning models. Moreover, we design a subject-specific training process to assist the model in fitting the data pattern of different patients or healthy people. Our method achieves a high accuracy of 73.8% in the cross-dataset classification. We conduct an extensive analysis to show the effectiveness of side information of ears in enhancing model performance and side-aware meta-learning in improving the quality of the learned features.

  View details for DOI 10.1109/TNSRE.2022.3201158

  View details for Web of Science ID 000848214700001

  View details for PubMedID 35998167

• Decoding SSVEP patterns from EEG via multivariate variational mode decomposition-informed canonical correlation analysis BIOMEDICAL SIGNAL PROCESSING AND CONTROL Chang, L., Wang, R., Zhang, Y. 2022; 71

  View details for DOI 10.1016/j.bspc.2021.103209

  View details for Web of Science ID 000710785900007

• Decoding Premovement Patterns with Task-Related Component Analysis COGNITIVE COMPUTATION Duan, F., Jia, H., Sun, Z., Zhang, K., Dai, Y., Zhang, Y. 2021; 13 (5): 1389-1405

  View details for DOI 10.1007/s12559-021-09941-7

  View details for Web of Science ID 000702804100001

• A Fast Approach to Removing Muscle Artifacts for EEG with Signal Serialization Based Ensemble Empirical Mode Decomposition. Entropy (Basel, Switzerland) Dai, Y., Duan, F., Feng, F., Sun, Z., Zhang, Y., Caiafa, C. F., Marti-Puig, P., Solé-Casals, J. 2021; 23 (9)

  Abstract

  An electroencephalogram (EEG) is an electrophysiological signal reflecting the functional state of the brain. As the control signal of the brain-computer interface (BCI), EEG may build a bridge between humans and computers to improve the life quality for patients with movement disorders. The collected EEG signals are extremely susceptible to the contamination of electromyography (EMG) artifacts, affecting their original characteristics. Therefore, EEG denoising is an essential preprocessing step in any BCI system. Previous studies have confirmed that the combination of ensemble empirical mode decomposition (EEMD) and canonical correlation analysis (CCA) can effectively suppress EMG artifacts. However, the time-consuming iterative process of EEMD may limit the application of the EEMD-CCA method in real-time monitoring of BCI. Compared with the existing EEMD, the recently proposed signal serialization based EEMD (sEEMD) is a good choice to provide effective signal analysis and fast mode decomposition. In this study, an EMG denoising method based on sEEMD and CCA is discussed. All of the analyses are carried out on semi-simulated data. The results show that, in terms of frequency and amplitude, the intrinsic mode functions (IMFs) decomposed by sEEMD are consistent with the IMFs obtained by EEMD. There is no significant difference in the ability to separate EMG artifacts from EEG signals between the sEEMD-CCA method and the EEMD-CCA method (p > 0.05). Even in the case of heavy contamination (signal-to-noise ratio is less than 2 dB), the relative root mean squared error is about 0.3, and the average correlation coefficient remains above 0.9. The running speed of the sEEMD-CCA method to remove EMG artifacts is significantly improved in comparison with that of EEMD-CCA method (p < 0.05). The running time of the sEEMD-CCA method for three lengths of semi-simulated data is shortened by more than 50%. This indicates that sEEMD-CCA is a promising tool for EMG artifact removal in real-time BCI systems.

  View details for DOI 10.3390/e23091170

  View details for PubMedID 34573795

  View details for PubMedCentralID PMC8465074

• Filter Bank-Driven Multivariate Synchronization Index for Training-Free SSVEP BCI. IEEE transactions on neural systems and rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society Qin, K., Wang, R., Zhang, Y. 2021; 29: 934-943

  Abstract

  In recent years, multivariate synchronization index (MSI) algorithm, as a novel frequency detection method, has attracted increasing attentions in the study of brain-computer interfaces (BCIs) based on steady state visual evoked potential (SSVEP). However, MSI algorithm is hard to fully exploit SSVEP-related harmonic components in the electroencephalogram (EEG), which limits the application of MSI algorithm in BCI systems. In this paper, we propose a novel filter bank-driven MSI algorithm (FBMSI) to overcome the limitation and further improve the accuracy of SSVEP recognition. We evaluate the efficacy of the FBMSI method by developing a 6-command SSVEP-NAO robot system with extensive experimental analyses. An offline experimental study is first performed with EEG collected from nine subjects to investigate the effects of varying parameters on the model performance. Offline results show that the proposed method has achieved a stable improvement effect. We further conduct an online experiment with six subjects to assess the efficacy of the developed FBMSI algorithm in a real-time BCI application. The online experimental results show that the FBMSI algorithm yields a promising average accuracy of 83.56% using a data length of even only one second, which was 12.26% higher than the standard MSI algorithm. These extensive experimental results confirmed the effectiveness of the FBMSI algorithm in SSVEP recognition and demonstrated its potential application in the development of improved BCI systems.

  View details for DOI 10.1109/TNSRE.2021.3073165

  View details for PubMedID 33852389

• Momentum contrastive learning for few-shot COVID-19 diagnosis from chest CT images. Pattern recognition Chen, X., Yao, L., Zhou, T., Dong, J., Zhang, Y. 2021; 113: 107826

  Abstract

  The current pandemic, caused by the outbreak of a novel coronavirus (COVID-19) in December 2019, has led to a global emergency that has significantly impacted economies, healthcare systems and personal wellbeing all around the world. Controlling the rapidly evolving disease requires highly sensitive and specific diagnostics. While RT-PCR is the most commonly used, it can take up to eight hours, and requires significant effort from healthcare professionals. As such, there is a critical need for a quick and automatic diagnostic system. Diagnosis from chest CT images is a promising direction. However, current studies are limited by the lack of sufficient training samples, as acquiring annotated CT images is time-consuming. To this end, we propose a new deep learning algorithm for the automated diagnosis of COVID-19, which only requires a few samples for training. Specifically, we use contrastive learning to train an encoder which can capture expressive feature representations on large and publicly available lung datasets and adopt the prototypical network for classification. We validate the efficacy of the proposed model in comparison with other competing methods on two publicly available and annotated COVID-19 CT datasets. Our results demonstrate the superior performance of our model for the accurate diagnosis of COVID-19 based on chest CT images.

  View details for DOI 10.1016/j.patcog.2021.107826

  View details for PubMedID 33518813

  View details for PubMedCentralID PMC7833525

• Multiview Feature Learning With Multiatlas-Based Functional Connectivity Networks for MCI Diagnosis. IEEE transactions on cybernetics Zhang, Y., Zhang, H., Adeli, E., Chen, X., Liu, M., Shen, D. 2020; PP

  Abstract

  Functional connectivity (FC) networks built from resting-state functional magnetic resonance imaging (rs-fMRI) has shown promising results for the diagnosis of Alzheimer's disease and its prodromal stage, that is, mild cognitive impairment (MCI). FC is usually estimated as a temporal correlation of regional mean rs-fMRI signals between any pair of brain regions, and these regions are traditionally parcellated with a particular brain atlas. Most existing studies have adopted a predefined brain atlas for all subjects. However, the constructed FC networks inevitably ignore the potentially important subject-specific information, particularly, the subject-specific brain parcellation. Similar to the drawback of the ``single view'' (versus the ``multiview'' learning) in medical image-based classification, FC networks constructed based on a single atlas may not be sufficient to reveal the underlying complicated differences between normal controls and disease-affected patients due to the potential bias from that particular atlas. In this study, we propose a multiview feature learning method with multiatlas-based FC networks to improve MCI diagnosis. Specifically, a three-step transformation is implemented to generate multiple individually specified atlases from the standard automated anatomical labeling template, from which a set of atlas exemplars is selected. Multiple FC networks are constructed based on these preselected atlas exemplars, providing multiple views of the FC network-based feature representations for each subject. We then devise a multitask learning algorithm for joint feature selection from the constructed multiple FC networks. The selected features are jointly fed into a support vector machine classifier for multiatlas-based MCI diagnosis. Extensive experimental comparisons are carried out between the proposed method and other competing approaches, including the traditional single-atlas-based method. The results indicate that our method significantly improves the MCI classification, demonstrating its promise in the brain connectome-based individualized diagnosis of brain diseases.

  View details for DOI 10.1109/TCYB.2020.3016953

  View details for PubMedID 33306476

• Multi-View Multi-Scale Optimization of Feature Representation for EEG Classification Improvement. IEEE transactions on neural systems and rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society Jiao, Y., Zhou, T., Yao, L., Zhou, G., Wang, X., Zhang, Y. 2020; 28 (12): 2589-2597

  Abstract

  Effectively extracting common space pattern (CSP) features from motor imagery (MI) EEG signals is often highly dependent on the filter band selection. At the same time, optimizing the EEG channel combinations is another key issue that substantially affects the SMR feature representations. Although numerous algorithms have been developed to find channels that record important characteristics of MI, most of them select channels in a cumbersome way with low computational efficiency, thereby limiting the practicality of MI-based BCI systems. In this study, we propose the multi-scale optimization (MSO) of spatial patterns, optimizing filter bands over multiple channel sets within CSPs to further improve the performance of MI-based BCI. Specifically, several channel subsets are first heuristically predefined, and then raw EEG data specific to each of these subsets bandpass-filtered at the overlap between a set of filter bands. Further, instead of solving learning problems for each channel subset independently, we propose a multi-view learning based sparse optimization to jointly extract robust CSP features with L2,1 -norm regularization, aiming to capture the shared salient information across multiple related spatial patterns for enhanced classification performance. A support vector machine (SVM) classifier is then trained on these optimized EEG features for accurate recognition of MI tasks. Experimental results on three public EEG datasets validate the effectiveness of MSO compared to several other competing methods and their variants. These superior experimental results demonstrate that the proposed MSO method has promising potential in MI-based BCIs.

  View details for DOI 10.1109/TNSRE.2020.3040984

  View details for PubMedID 33245696

• A survey on deep learning-based non-invasive brain signals: recent advances and new frontiers. Journal of neural engineering Zhang, X., Yao, L., Wang, X., Monaghan, J. J., Mcalpine, D., Zhang, Y. 2020

  Abstract

  Brain signals refer to the biometric information collected from the human brain. The research on brain signals aims to discover the underlying neurological or physical status of the individuals by signal decoding. The emerging deep learning techniques have improved the study of brain signals significantly in recent years. In this work, we first present a taxonomy of non-invasive brain signals and the basics of deep learning algorithms. Then, we provide a comprehensive survey of the frontiers of applying deep learning for non-invasive brain signals analysis, by summarizing a large number of recent publications. Moreover, upon the deep learning-powered brain signal studies, we report the potential real-world applications which benefit not only disabled people but also normal individuals. Finally, we discuss the opening challenges and future directions.

  View details for DOI 10.1088/1741-2552/abc902

  View details for PubMedID 33171452

• Adversarial Representation Learning for Robust Patient-Independent Epileptic Seizure Detection IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS Zhang, X., Yao, L., Dong, M., Liu, Z., Zhang, Y., Li, Y. 2020; 24 (10): 2852–59

  Abstract

  Epilepsy is a chronic neurological disorder characterized by the occurrence of spontaneous seizures, which affects about one percent of the worlds population. Most of the current seizure detection approaches strongly rely on patient history records and thus fail in the patient-independent situation of detecting the new patients. To overcome such limitation, we propose a robust and explainable epileptic seizure detection model that effectively learns from seizure states while eliminates the inter-patient noises. A complex deep neural network model is proposed to learn the pure seizure-specific representation from the raw non-invasive electroencephalography (EEG) signals through adversarial training. Furthermore, to enhance the explainability, we develop an attention mechanism to automatically learn the importance of each EEG channels in the seizure diagnosis procedure. The proposed approach is evaluated over the Temple University Hospital EEG (TUH EEG) database. The experimental results illustrate that our model outperforms the competitive state-of-the-art baselines with low latency. Moreover, the designed attention mechanism is demonstrated ables to provide fine-grained information for pathological analysis. We propose an effective and efficient patient-independent diagnosis approach of epileptic seizure based on raw EEG signals without manually feature engineering, which is a step toward the development of large-scale deployment for real-life use.

  View details for DOI 10.1109/JBHI.2020.2971610

  View details for Web of Science ID 000576429900014

  View details for PubMedID 32071011

• EEG classification using sparse Bayesian extreme learning machine for brain-computer interface NEURAL COMPUTING & APPLICATIONS Jin, Z., Zhou, G., Gao, D., Zhang, Y. 2020; 32 (11): 6601–9

  View details for DOI 10.1007/s00521-018-3735-3

  View details for Web of Science ID 000536371900018

• Verbal Memory Abilities and EEG Resting State Connectivity Predict Therapy Outcome in Veterans With PTSD Naparstek, S., Narayan, M., Maron-Katz, A., Zhang, Y., Etkin, A. ELSEVIER SCIENCE INC. 2020: S461

  View details for Web of Science ID 000535308201373

• An electroencephalographic signature predicts antidepressant response in major depression. Nature biotechnology Wu, W., Zhang, Y., Jiang, J., Lucas, M. V., Fonzo, G. A., Rolle, C. E., Cooper, C., Chin-Fatt, C., Krepel, N., Cornelssen, C. A., Wright, R., Toll, R. T., Trivedi, H. M., Monuszko, K., Caudle, T. L., Sarhadi, K., Jha, M. K., Trombello, J. M., Deckersbach, T., Adams, P., McGrath, P. J., Weissman, M. M., Fava, M., Pizzagalli, D. A., Arns, M., Trivedi, M. H., Etkin, A. 2020

  Abstract

  Antidepressants are widely prescribed, but their efficacy relative to placebo is modest, in part because the clinical diagnosis of major depression encompasses biologically heterogeneous conditions. Here, we sought to identify a neurobiological signature of response to antidepressant treatment as compared to placebo. We designed a latent-space machine-learning algorithm tailored for resting-state electroencephalography (EEG) and applied it to data from the largest imaging-coupled, placebo-controlled antidepressant study (n=309). Symptom improvement was robustly predicted in a manner both specific for the antidepressant sertraline (versus placebo) and generalizable across different study sites and EEG equipment. This sertraline-predictive EEG signature generalized to two depression samples, wherein it reflected general antidepressant medication responsivity and related differentially to a repetitive transcranial magnetic stimulation treatment outcome. Furthermore, we found that the sertraline resting-state EEG signature indexed prefrontal neural responsivity, as measured by concurrent transcranial magnetic stimulation and EEG. Our findings advance the neurobiological understanding of antidepressant treatment through an EEG-tailored computational model and provide a clinical avenue for personalized treatment of depression.

  View details for DOI 10.1038/s41587-019-0397-3

  View details for PubMedID 32042166

• An Electroencephalography Connectomic Profile of Posttraumatic Stress Disorder. The American journal of psychiatry Toll, R. T., Wu, W., Naparstek, S., Zhang, Y., Narayan, M., Patenaude, B., De Los Angeles, C., Sarhadi, K., Anicetti, N., Longwell, P., Shpigel, E., Wright, R., Newman, J., Gonzalez, B., Hart, R., Mann, S., Abu-Amara, D., Sarhadi, K., Cornelssen, C., Marmar, C., Etkin, A. 2020: appiajp201918080911

  Abstract

  OBJECTIVE: The authors sought to identify brain regions whose frequency-specific, orthogonalized resting-state EEG power envelope connectivity differs between combat veterans with posttraumatic stress disorder (PTSD) and healthy combat-exposed veterans, and to determine the behavioral correlates of connectomic differences.METHODS: The authors first conducted a connectivity method validation study in healthy control subjects (N=36). They then conducted a two-site case-control study of veterans with and without PTSD who were deployed to Iraq and/or Afghanistan. Healthy individuals (N=95) and those meeting full or subthreshold criteria for PTSD (N=106) underwent 64-channel resting EEG (eyes open and closed), which was then source-localized and orthogonalized to mitigate effects of volume conduction. Correlation coefficients between band-limited source-space power envelopes of different regions of interest were then calculated and corrected for multiple comparisons. Post hoc correlations of connectomic abnormalities with clinical features and performance on cognitive tasks were conducted to investigate the relevance of the dysconnectivity findings.RESULTS: Seventy-four brain region connections were significantly reduced in PTSD (all in the eyes-open condition and predominantly using the theta carrier frequency). Underconnectivity of the orbital and anterior middle frontal gyri were most prominent. Performance differences in the digit span task mapped onto connectivity between 25 of the 74 brain region pairs, including within-network connections in the dorsal attention, frontoparietal control, and ventral attention networks.CONCLUSIONS: Robust PTSD-related abnormalities were evident in theta-band source-space orthogonalized power envelope connectivity, which furthermore related to cognitive deficits in these patients. These findings establish a clinically relevant connectomic profile of PTSD using a tool that facilitates the lower-cost clinical translation of network connectivity research.

  View details for DOI 10.1176/appi.ajp.2019.18080911

  View details for PubMedID 31964161

• An ERP-based BCI with peripheral stimuli: validation with ALS patients. Cognitive neurodynamics Miao, Y. n., Yin, E. n., Allison, B. Z., Zhang, Y. n., Chen, Y. n., Dong, Y. n., Wang, X. n., Hu, D. n., Chchocki, A. n., Jin, J. n. 2020; 14 (1): 21–33

  Abstract

  Many studies reported that ERP-based BCIs can provide communication for some people with amyotrophic lateral sclerosis (ALS). ERP-based BCIs often present characters within a matrix that occupies the center of the visual field. However, several studies have identified some concerns with the matrix-based approach. This approach may lead to fatigue and errors resulting from flashing adjacent stimuli, and is impractical for users who might want to use the BCI in tandem with other software or feedback in the center of the monitor. In this paper, we introduce and validate an alternate ERP-based BCI display approach. By presenting stimuli near the periphery of the display, we reduce the adjacency problem and leave the center of the display available for feedback or other applications. Two ERP-based display approaches were tested on 18 ALS patients to: (1) compare performance between a conventional matrix speller paradigm (Matrix-P, mean visual angle 6°) and a new speller paradigm with peripherally distributed stimuli (Peripheral-P, mean visual angle 8.8°); and (2) assess performance while spelling 42 characters online continuously, without a break. In the Peripheral-P condition, 12 subjects attained higher than 80% feedback accuracy during online performance, and 7 of these subjects obtained higher than 90% accuracy. The experimental results showed that the Peripheral-P condition yielded performance comparable to the conventional Matrix-P condition (p > 0.05) in accuracy and information transfer rate. This paper introduces a new display approach that leaves the center of the monitor open for feedback and/or other display elements, such as movies, games, art, or displays from other AAC software or conventional software tools.

  View details for DOI 10.1007/s11571-019-09541-0

  View details for PubMedID 32015765

  View details for PubMedCentralID PMC6974196

• Identification of psychiatric disorder subtypes from functional connectivity patterns in resting-state electroencephalography. Nature Biomedical Engineering Zhang, Y., Wu, W., Toll, R. T., Naparstek, S., Maron-katz, A., Watts, M., Gorddodn, J., Jeeong, J., Astolfi, L., Shpigel, E., Longwell, P., Sarhadi, k., El-Said, D., Li, Y., Cooper, C., Chin-Fatt, C., Arns, M., Goodkind, M. S., Trivedi, M. H., Marmar, C. R., Etkin, A. 2020

  View details for DOI 10.1038/s41551-020-00614-8

• A toolbox for brain network construction and classification (BrainNetClass). Human brain mapping Zhou, Z. n., Chen, X. n., Zhang, Y. n., Hu, D. n., Qiao, L. n., Yu, R. n., Yap, P. T., Pan, G. n., Zhang, H. n., Shen, D. n. 2020

  Abstract

  Brain functional network has been increasingly used in understanding brain functions and diseases. While many network construction methods have been proposed, the progress in the field still largely relies on static pairwise Pearson's correlation-based functional network and group-level comparisons. We introduce a "Brain Network Construction and Classification (BrainNetClass)" toolbox to promote more advanced brain network construction methods to the filed, including some state-of-the-art methods that were recently developed to capture complex and high-order interactions among brain regions. The toolbox also integrates a well-accepted and rigorous classification framework based on brain connectome features toward individualized disease diagnosis in a hope that the advanced network modeling could boost the subsequent classification. BrainNetClass is a MATLAB-based, open-source, cross-platform toolbox with both graphical user-friendly interfaces and a command line mode targeting cognitive neuroscientists and clinicians for promoting reliability, reproducibility, and interpretability of connectome-based, computer-aided diagnosis. It generates abundant classification-related results from network presentations to contributing features that have been largely ignored by most studies to grant users the ability of evaluating the disease diagnostic model and its robustness and generalizability. We demonstrate the effectiveness of the toolbox on real resting-state functional MRI datasets. BrainNetClass (v1.0) is available at https://github.com/zzstefan/BrainNetClass.

  View details for DOI 10.1002/hbm.24979

  View details for PubMedID 32163221

• A Novel EEMD-CCA Approach to Removing Muscle Artifacts for Pervasive EEG IEEE SENSORS JOURNAL Chen, X., Chen, Q., Zhang, Y., Wang, Z. 2019; 19 (19): 8420–31

  View details for DOI 10.1109/JSEN.2018.2872623

  View details for Web of Science ID 000487216200008

• Temporally Constrained Sparse Group Spatial Patterns for Motor Imagery BCI IEEE TRANSACTIONS ON CYBERNETICS Zhang, Y., Nam, C. S., Zhou, G., Jin, J., Wang, X., Cichocki, A. 2019; 49 (9): 3322–32

  Abstract

  Common spatial pattern (CSP)-based spatial filtering has been most popularly applied to electroencephalogram (EEG) feature extraction for motor imagery (MI) classification in brain-computer interface (BCI) application. The effectiveness of CSP is highly affected by the frequency band and time window of EEG segments. Although numerous algorithms have been designed to optimize the spectral bands of CSP, most of them selected the time window in a heuristic way. This is likely to result in a suboptimal feature extraction since the time period when the brain responses to the mental tasks occurs may not be accurately detected. In this paper, we propose a novel algorithm, namely temporally constrained sparse group spatial pattern (TSGSP), for the simultaneous optimization of filter bands and time window within CSP to further boost classification accuracy of MI EEG. Specifically, spectrum-specific signals are first derived by bandpass filtering from raw EEG data at a set of overlapping filter bands. Each of the spectrum-specific signals is further segmented into multiple subseries using sliding window approach. We then devise a joint sparse optimization of filter bands and time windows with temporal smoothness constraint to extract robust CSP features under a multitask learning framework. A linear support vector machine classifier is trained on the optimized EEG features to accurately identify the MI tasks. An experimental study is implemented on three public EEG datasets (BCI Competition III dataset IIIa, BCI Competition IV datasets IIa, and BCI Competition IV dataset IIb) to validate the effectiveness of TSGSP in comparison to several other competing methods. Superior classification performance (averaged accuracies are 88.5%, 83.3%, and 84.3% for the three datasets, respectively) based on the experimental results confirms that the proposed algorithm is a promising candidate for performance improvement of MI-based BCIs.

  View details for DOI 10.1109/TCYB.2018.2841847

  View details for Web of Science ID 000470988800009

  View details for PubMedID 29994667

• Regularized Group Sparse Discriminant Analysis for P300-Based Brain-Computer Interface INTERNATIONAL JOURNAL OF NEURAL SYSTEMS Wu, Q., Zhang, Y., Liu, J., Sun, J., Cichocki, A., Gao, F. 2019; 29 (6)

  View details for DOI 10.1142/S0129065719500023

  View details for Web of Science ID 000477829200005

• Strength and similarity guided group-level brain functional network construction for MCI diagnosis PATTERN RECOGNITION Zhang, Y., Zhang, H., Chen, X., Liu, M., Zhu, X., Lee, S., Shen, D. 2019; 88: 421–30

  View details for DOI 10.1016/j.patcog.2018.12.001

  View details for Web of Science ID 000457666900033

• Strength and Similarity Guided Group-level Brain Functional Network Construction for MCI Diagnosis. Pattern recognition Zhang, Y., Zhang, H., Chen, X., Liu, M., Zhu, X., Lee, S. W., Shen, D. 2019; 88: 421-430

  Abstract

  Sparse representation-based brain functional network modeling often results in large inter-subject variability in the network structure. This could reduce the statistical power in group comparison, or even deteriorate the generalization capability of the individualized diagnosis of brain diseases. Although group sparse representation (GSR) can alleviate such a limitation by increasing network similarity across subjects, it could, in turn, fail in providing satisfactory separability between the subjects from different groups (e.g., patients vs. controls). In this study, we propose to integrate individual functional connectivity (FC) information into the GSR-based network construction framework to achieve higher between-group separability while maintaining the merit of within-group consistency. Our method was based on an observation that the subjects from the same group have generally more similar FC patterns than those from different groups. To this end, we propose our new method, namely "strength and similarity guided GSR (SSGSR)", which exploits both BOLD signal temporal correlation-based "low-order" FC (LOFC) and inter-subject LOFC-profile similarity-based "high-order" FC (HOFC) as two priors to jointly guide the GSR-based network modeling. Extensive experimental comparisons are carried out, with the rs-fMRI data from mild cognitive impairment (MCI) subjects and healthy controls, between the proposed algorithm and other state-of-the-art brain network modeling approaches. Individualized MCI identification results show that our method could achieve a balance between the individually consistent brain functional network construction and the adequately maintained inter-group brain functional network distinctions, thus leading to a more accurate classification result. Our method also provides a promising and generalized solution for the future connectome-based individualized diagnosis of brain disease.

  View details for DOI 10.1016/j.patcog.2018.12.001

  View details for PubMedID 31579344

  View details for PubMedCentralID PMC6774624

• Sparse Group Representation Model for Motor Imagery EEG Classification IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS Jiao, Y., Zhang, Y., Chen, X., Yin, E., Jin, J., Wang, X., Cichocki, A. 2019; 23 (2): 631–41

  Abstract

  A potential limitation of a motor imagery (MI) based brain-computer interface (BCI) is that it usually requires a relatively long time to record sufficient electroencephalogram (EEG) data for robust classifier training. The calibration burden during data acquisition phase will most probably cause a subject to be reluctant to use a BCI system. To alleviate this issue, we propose a novel sparse group representation model (SGRM) for improving the efficiency of MI-based BCI by exploiting the intersubject information. Specifically, preceded by feature extraction using common spatial pattern, a composite dictionary matrix is constructed with training samples from both the target subject and other subjects. By explicitly exploiting within-group sparse and group-wise sparse constraints, the most compact representation of a test sample of the target subject is then estimated as a linear combination of columns in the dictionary matrix. Classification is implemented by calculating the class-specific representation residual based on the significant training samples corresponding to the nonzero representation coefficients. Accordingly, the proposed SGRM method effectively reduces the required training samples from the target subject due to auxiliary data available from other subjects. With two public EEG data sets, extensive experimental comparisons are carried out between SGRM and other state-of-the-art approaches. Superior classification performance of our method using 40 trials of the target subject for model calibration (Averaged accuracy = 78.2%, Kappa = 0.57 and Averaged accuracy = 77.7%, Kappa = 0.55 for the two data sets, respectively) indicates its promising potential for improving the practicality of MI-based BCI.

  View details for DOI 10.1109/JBHI.2018.2832538

  View details for Web of Science ID 000460666400019

  View details for PubMedID 29994055

• Effect of a combination of flip and zooming stimuli on the performance of a visual brain-computer interface for spelling. Biomedizinische Technik. Biomedical engineering Cheng, J., Jin, J., Daly, I., Zhang, Y., Wang, B., Wang, X., Cichocki, A. 2019; 64 (1): 29–38

  Abstract

  Brain-computer interface (BCI) systems can allow their users to communicate with the external world by recognizing intention directly from their brain activity without the assistance of the peripheral motor nervous system. The P300-speller is one of the most widely used visual BCI applications. In previous studies, a flip stimulus (rotating the background area of the character) that was based on apparent motion, suffered from less refractory effects. However, its performance was not improved significantly. In addition, a presentation paradigm that used a "zooming" action (changing the size of the symbol) has been shown to evoke relatively higher P300 amplitudes and obtain a better BCI performance. To extend this method of stimuli presentation within a BCI and, consequently, to improve BCI performance, we present a new paradigm combining both the flip stimulus with a zooming action. This new presentation modality allowed BCI users to focus their attention more easily. We investigated whether such an action could combine the advantages of both types of stimuli presentation to bring a significant improvement in performance compared to the conventional flip stimulus. The experimental results showed that the proposed paradigm could obtain significantly higher classification accuracies and bit rates than the conventional flip paradigm (p<0.01).

  View details for PubMedID 29432199

• Regularized Group Sparse Discriminant Analysis for P300-Based Brain-Computer Interface. International journal of neural systems Wu, Q., Zhang, Y., Liu, J., Sun, J., Cichocki, A., Gao, F. 2019: 1950002

  Abstract

  Event-related potentials (ERPs) especially P300 are popular effective features for brain-computer interface (BCI) systems based on electroencephalography (EEG). Traditional ERP-based BCI systems may perform poorly for small training samples, i.e. the undersampling problem. In this study, the ERP classification problem was investigated, in particular, the ERP classification in the high-dimensional setting with the number of features larger than the number of samples was studied. A flexible group sparse discriminative analysis algorithm based on Moreau-Yosida regularization was proposed for alleviating the undersampling problem. An optimization problem with the group sparse criterion was presented, and the optimal solution was proposed by using the regularized optimal scoring method. During the alternating iteration procedure, the feature selection and classification were performed simultaneously. Two P300-based BCI datasets were used to evaluate our proposed new method and compare it with existing standard methods. The experimental results indicated that the features extracted via our proposed method are efficient and provide an overall better P300 classification accuracy compared with several state-of-the-art methods.

  View details for PubMedID 30880525

• Hierarchical feature fusion framework for frequency recognition in SSVEP-based BCIs. Neural networks : the official journal of the International Neural Network Society Zhang, Y. n., Yin, E. n., Li, F. n., Zhang, Y. n., Guo, D. n., Yao, D. n., Xu, P. n. 2019; 119: 1–9

  Abstract

  Effective frequency recognition algorithms are critical in steady-state visual evoked potential (SSVEP) based brain-computer interfaces (BCIs). In this study, we present a hierarchical feature fusion framework which can be used to design high-performance frequency recognition methods. The proposed framework includes two primary techniques for fusing features: spatial dimension fusion (SD) and frequency dimension fusion (FD). Both SD and FD fusions are obtained using a weighted strategy with a nonlinear function. To assess our novel methods, we used the correlated component analysis (CORRCA) method to investigate the efficiency and effectiveness of the proposed framework. Experimental results were obtained from a benchmark dataset of thirty-five subjects and indicate that the extended CORRCA method used within the framework significantly outperforms the original CORCCA method. Accordingly, the proposed framework holds promise to enhance the performance of frequency recognition methods in SSVEP-based BCIs.

  View details for DOI 10.1016/j.neunet.2019.07.007

  View details for PubMedID 31376634

• Brain regulation of emotional conflict predicts antidepressant treatment response for depression. Nature human behaviour Fonzo, G. A., Etkin, A. n., Zhang, Y. n., Wu, W. n., Cooper, C. n., Chin-Fatt, C. n., Jha, M. K., Trombello, J. n., Deckersbach, T. n., Adams, P. n., McInnis, M. n., McGrath, P. J., Weissman, M. M., Fava, M. n., Trivedi, M. H. 2019

  Abstract

  The efficacy of antidepressant treatment for depression is controversial due to the only modest superiority demonstrated over placebo. However, neurobiological heterogeneity within depression may limit overall antidepressant efficacy. We sought to identify a neurobiological phenotype responsive to antidepressant treatment by testing pretreatment brain activation during response to, and regulation of, emotional conflict as a moderator of the clinical benefit of the antidepressant sertraline versus placebo. Using neuroimaging data from a large randomized controlled trial, we found widespread moderation of clinical benefits by brain activity during regulation of emotional conflict, in which greater downregulation of conflict-responsive regions predicted better sertraline outcomes. Treatment-predictive machine learning using brain metrics outperformed a model trained on clinical and demographic variables. Our findings demonstrate that antidepressant response is predicted by brain activity underlying a key self-regulatory emotional capacity. Leveraging brain-based measures in psychiatry will forge a path toward better treatment personalization, refined mechanistic insights and improved outcomes.

  View details for DOI 10.1038/s41562-019-0732-1

  View details for PubMedID 31548678

• Individual Patterns of Abnormality in Resting-State Functional Connectivity Reveal Two Data-Driven PTSD Subgroups. The American journal of psychiatry Maron-Katz, A. n., Zhang, Y. n., Narayan, M. n., Wu, W. n., Toll, R. T., Naparstek, S. n., De Los Angeles, C. n., Longwell, P. n., Shpigel, E. n., Newman, J. n., Abu-Amara, D. n., Marmar, C. n., Etkin, A. n. 2019: appiajp201919010060

  Abstract

  A major challenge in understanding and treating posttraumatic stress disorder (PTSD) is its clinical heterogeneity, which is likely determined by various neurobiological perturbations. This heterogeneity likely also reduces the effectiveness of standard group comparison approaches. The authors tested whether a statistical approach aimed at identifying individual-level neuroimaging abnormalities that are more prevalent in case subjects than in control subjects could reveal new clinically meaningful insights into the heterogeneity of PTSD.Resting-state functional MRI data were recorded from 87 unmedicated PTSD case subjects and 105 war zone-exposed healthy control subjects. Abnormalities were modeled using tolerance intervals, which referenced the distribution of healthy control subjects as the "normative population." Out-of-norm functional connectivity values were examined for enrichment in cases and then used in a clustering analysis to identify biologically defined PTSD subgroups based on their abnormality profiles.The authors identified two subgroups among PTSD cases, each with a distinct pattern of functional connectivity abnormalities with respect to healthy control subjects. Subgroups differed clinically on levels of reexperiencing symptoms and improved case-control discriminability and were detectable using independently recorded resting-state EEG data.The results provide proof of concept for the utility of abnormality-based approaches for studying heterogeneity within clinical populations. Such approaches, applied not only to neuroimaging data, may allow detection of subpopulations with distinct biological signatures so that further clinical and mechanistic investigations can be focused on more biologically homogeneous subgroups.

  View details for DOI 10.1176/appi.ajp.2019.19010060

  View details for PubMedID 31838870

• Inter-modality Dependence Induced Data Recovery for MCI Conversion Prediction Zhou, T., Thung, K., Zhang, Y., Fu, H., Shen, J., Shen, D., Shao, L., Shen, D., Liu, T., Peters, T. M., Staib, L. H., Essert, C., Zhou, S., Yap, P. T., Khan, A. SPRINGER INTERNATIONAL PUBLISHING AG. 2019: 186–95

  View details for DOI 10.1007/978-3-030-32251-9_21

  View details for Web of Science ID 000548735900021

• Correlated Component Analysis for Enhancing the Performance of SSVEP-Based Brain-Computer Interface (vol 26, pg 948, 2018) IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING Zhang, Y., Guo, D., Li, F., Yin, E., Zhang, Y., Li, P., Zhao, Q., Tanaka, T., Yao, D., Xu, P., Nakanishi, M. 2018; 26 (8): 1645–46

  Abstract

  In the above paper [1], a method has been proposed to use the correlated component analysis (CORCA) to learn spatial filters with multiple blocks of individual training data for steady-state visual evoked potential (SSVEP)-based brain-computer interface (BCI) scenario. In order to evaluate the performance of CORCA, the task-related component analysis (TRCA)-based method was used as a baseline method [2]. For a fair and convincing comparison, the MATLAB codes on the website (https://github.com/mnakanishi/TRCA-SSVEP) for implementing TRCA method provided by Dr. Masaki Nakanishi, the first author of [2], were used to take the role of the TRCA method. At that time, the proposed CORCA-based method outperforms the TRCA-based method [1].

  View details for DOI 10.1109/TNSRE.2018.2851318

  View details for Web of Science ID 000441426400017

  View details for PubMedID 30102598

• Removal of muscle artefacts from few-channel EEG recordings based on multivariate empirical mode decomposition and independent vector analysis ELECTRONICS LETTERS Xu, X., Chen, X., Zhang, Y. 2018; 54 (14): 866–67

  View details for DOI 10.1049/el.2018.0191

  View details for Web of Science ID 000437171500005

• Two-Stage Frequency Recognition Method Based on Correlated Component Analysis for SSVEP-Based BCI. IEEE transactions on neural systems and rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society Zhang, Y., Yin, E., Li, F., Zhang, Y., Tanaka, T., Zhao, Q., Cui, Y., Xu, P., Yao, D., Guo, D. 2018; 26 (7): 1314–23

  Abstract

  A canonical correlation analysis (CCA) is a state-of-the-art method for frequency recognition in steady-state visual evoked potential (SSVEP)-based brain-computer interface (BCI) systems. Various extended methods have been developed, and among such methods, a combination method of CCA and individual-template-based CCA has achieved the best performance. However, the CCA requires the canonical vectors to be orthogonal, which may not be a reasonable assumption for the EEG analysis. In this paper, we propose using the correlated component analysis (CORRCA) rather than CCA to implement frequency recognition. CORRCA can relax the constraint of canonical vectors in CCA and generate the same projection vector for two multichannel EEG signals. Furthermore, we propose a two-stage method based on the basic CORRCA method (termed TSCORRCA). Evaluated on a benchmark data set of 35 subjects, the experimental results demonstrate that CORRCA significantly outperformed CCA, and TSCORRCA obtained the best performance among the compared methods. This paper demonstrates that CORRCA-based methods have a great potential for implementing high-performance SSVEP-based BCI systems.

  View details for DOI 10.1109/TNSRE.2018.2848222

  View details for PubMedID 29985141

• PTSD Subtype Identification Based on Resting-State EEG Functional Connectivity Biomarkers Zhang, Y., Toll, R., Wu, W., Longwell, P., Shpigel, E., Abu Amara, D., Gonzalez, B., Mann, S., Hart, R., Marmar, C., Etkin, A. ELSEVIER SCIENCE INC. 2018: S141

  View details for Web of Science ID 000432466300346

• Correlated Component Analysis for Enhancing the Performance of SSVEP-Based Brain-Computer Interface IEEE TRANSACTIONS ON NEURAL SYSTEMS AND REHABILITATION ENGINEERING Zhang, Y., Guo, D., Li, F., Yin, E., Zhang, Y., Li, P., Zhao, Q., Tanaka, T., Yao, D., Xu, P. 2018; 26 (5): 948–56

  Abstract

  A new method for steady-state visual evoked potentials (SSVEPs) frequency recognition is proposed to enhance the performance of SSVEP-based brain-computer interface (BCI). Correlated component analysis (CORCA) is introduced, which originally was designed to find linear combinations of electrodes that are consistent across subjects and maximally correlated between them. We propose a CORCA algorithm to learn spatial filters with multiple blocks of individual training data for SSVEP-based BCI scenario. The spatial filters are used to remove background noises by combining the multichannel electroencephalogram signals. We conduct a comparison between the proposed CORCA-based and the task-related component analysis (TRCA) based methods using a 40-class SSVEP benchmark data set recorded from 35 subjects. Our experimental study validates the efficiency of the CORCA-based method, and the extensive comparison results indicate that the CORCA-based method significantly outperforms the TRCA-based method. Superior performance demonstrates that the proposed method holds the promising potential to achieve satisfactory performance for SSVEP-based BCI with a large number of targets.

  View details for DOI 10.1109/TNSRE.2018.2826541

  View details for Web of Science ID 000432011000003

  View details for PubMedID 29752229

• Naturalistic Clinical Monitoring of rTMS-Induced Plasticity With TMS-EEG Keller, C., Wu, W., Sarhadi, K., Zhang, Y., Kerwin, L., Bhati, M., Etkin, A. ELSEVIER SCIENCE INC. 2018: S195

  View details for Web of Science ID 000432466300487

• Multi-kernel extreme learning machine for EEG classification in brain-computer interfaces EXPERT SYSTEMS WITH APPLICATIONS Zhang, Y., Wang, Y., Zhou, G., Jin, J., Wang, B., Wang, X., Cichocki, A. 2018; 96: 302-310

  View details for DOI 10.1016/j.eswa.2017.12.015

  View details for Web of Science ID 000424176900023

• Usage of drip drops as stimuli in an auditory P300 BCI paradigm. Cognitive neurodynamics Huang, M., Jin, J., Zhang, Y., Hu, D., Wang, X. 2018; 12 (1): 85-94

  Abstract

  Recently, many auditory BCIs are using beeps as auditory stimuli, while beeps sound unnatural and unpleasant for some people. It is proved that natural sounds make people feel comfortable, decrease fatigue, and improve the performance of auditory BCI systems. Drip drop is a kind of natural sounds that makes humans feel relaxed and comfortable. In this work, three kinds of drip drops were used as stimuli in an auditory-based BCI system to improve the user-friendness of the system. This study explored whether drip drops could be used as stimuli in the auditory BCI system. The auditory BCI paradigm with drip-drop stimuli, which was called the drip-drop paradigm (DP), was compared with the auditory paradigm with beep stimuli, also known as the beep paradigm (BP), in items of event-related potential amplitudes, online accuracies and scores on the likability and difficulty to demonstrate the advantages of DP. DP obtained significantly higher online accuracy and information transfer rate than the BP (p < 0.05, Wilcoxon signed test; p < 0.05, Wilcoxon signed test). Besides, DP obtained higher scores on the likability with no significant difference on the difficulty (p < 0.05, Wilcoxon signed test). The results showed that the drip drops were reliable acoustic materials as stimuli in an auditory BCI system.

  View details for DOI 10.1007/s11571-017-9456-y

  View details for PubMedID 29435089

  View details for PubMedCentralID PMC5801279

• Exploiting Convolutional Neural Networks With Deeply Local Description for Remote Sensing Image Classification IEEE ACCESS Liu, N., Wan, L., Zhang, Y., Zhou, T., Huo, H., Fang, T. 2018; 6: 11215–28

  View details for DOI 10.1109/ACCESS.2018.2798799

  View details for Web of Science ID 000427914200001

• Test-Retest Reliability of "High-Order" Functional Connectivity in Young Healthy Adults. Frontiers in neuroscience Zhang, H., Chen, X., Zhang, Y., Shen, D. 2017; 11: 439

  Abstract

  Functional connectivity (FC) has become a leading method for resting-state functional magnetic resonance imaging (rs-fMRI) analysis. However, the majority of the previous studies utilized pairwise, temporal synchronization-based FC. Recently, high-order FC (HOFC) methods were proposed with the idea of computing "correlation of correlations" to capture high-level, more complex associations among the brain regions. There are two types of HOFC. The first type is topographical profile similarity-based HOFC (tHOFC) and its variant, associated HOFC (aHOFC), for capturing different levels of HOFC. Instead of measuring the similarity of the original rs-fMRI signals with the traditional FC (low-order FC, or LOFC), tHOFC measures the similarity of LOFC profiles (i.e., a set of LOFC values between a region and all other regions) between each pair of brain regions. The second type is dynamics-based HOFC (dHOFC) which defines the quadruple relationship among every four brain regions by first calculating two pairwise dynamic LOFC "time series" and then measuring their temporal synchronization (i.e., temporal correlation of the LOFC fluctuations, not the BOLD fluctuations). Applications have shown the superiority of HOFC in both disease biomarker detection and individualized diagnosis than LOFC. However, no study has been carried out for the assessment of test-retest reliability of different HOFC metrics. In this paper, we systematically evaluate the reliability of the two types of HOFC methods using test-retest rs-fMRI data from 25 (12 females, age 24.48 ± 2.55 years) young healthy adults with seven repeated scans (with interval = 3-8 days). We found that all HOFC metrics have satisfactory reliability, specifically (1) fair-to-good for tHOFC and aHOFC, and (2) fair-to-moderate for dHOFC with relatively strong connectivity strength. We further give an in-depth analysis of the biological meanings of each HOFC metric and highlight their differences compared to the LOFC from the aspects of cross-level information exchanges, within-/between-network connectivity, and modulatory connectivity. In addition, how the dynamic analysis parameter (i.e., sliding window length) affects dHOFC reliability is also investigated. Our study reveals unique functional associations characterized by the HOFC metrics. Guidance and recommendations for future applications and clinical research using HOFC are provided. This study has made a further step toward unveiling more complex human brain connectome.

  View details for DOI 10.3389/fnins.2017.00439

  View details for PubMedID 28824362

  View details for PubMedCentralID PMC5539178

• Bayesian Nonnegative CP Decomposition-Based Feature Extraction Algorithm for Drowsiness Detection. IEEE transactions on neural systems and rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society Qian, D., Wang, B., Qing, X., Zhang, T., Zhang, Y., Wang, X., Nakamura, M. 2017; 25 (8): 1297-1308

  Abstract

  Daytime short nap involves physiological processes, such as alertness, drowsiness and sleep. The study of the relationship between drowsiness and nap based on physiological signals is a great way to have a better understanding of the periodical rhymes of physiological states. A model of Bayesian nonnegative CP decomposition (BNCPD) was proposed to extract common multiway features from the group-level electroencephalogram (EEG) signals. As an extension of the nonnegative CP decomposition, the BNCPD model involves prior distributions of factor matrices, while the underlying CP rank could be determined automatically based on a Bayesian nonparametric approach. In terms of computational speed, variational inference was applied to approximate the posterior distributions of unknowns. Extensive simulations on the synthetic data illustrated the capability of our model to recover the true CP rank. As a real-world application, the performance of drowsiness detection during daytime short nap by using the BNCPD-based features was compared with that of other traditional feature extraction methods. Experimental results indicated that the BNCPD model outperformed other methods for feature extraction in terms of two evaluation metrics, as well as different parameter settings. Our approach is likely to be a useful tool for automatic CP rank determination and offering a plausible multiway physiological information of individual states.

  View details for DOI 10.1109/TNSRE.2016.2618902

  View details for PubMedID 27775525

• Hybrid High-order Functional Connectivity Networks Using Resting-state Functional MRI for Mild Cognitive Impairment Diagnosis. Scientific reports Zhang, Y., Zhang, H., Chen, X., Lee, S. W., Shen, D. 2017; 7 (1): 6530

  Abstract

  Conventional functional connectivity (FC), referred to as low-order FC, estimates temporal correlation of the resting-state functional magnetic resonance imaging (rs-fMRI) time series between any pair of brain regions, simply ignoring the potentially high-level relationship among these brain regions. A high-order FC based on "correlation's correlation" has emerged as a new approach for abnormality detection of brain disease. However, separate construction of the low- and high-order FC networks overlooks information exchange between the two FC levels. Such a higher-level relationship could be more important for brain diseases study. In this paper, we propose a novel framework, namely "hybrid high-order FC networks" by exploiting the higher-level dynamic interaction among brain regions for early mild cognitive impairment (eMCI) diagnosis. For each sliding window-based rs-fMRI sub-series, we construct a whole-brain associated high-order network, by estimating the correlations between the topographical information of the high-order FC sub-network from one brain region and that of the low-order FC sub-network from another brain region. With multi-kernel learning, complementary features from multiple time-varying FC networks constructed at different levels are fused for eMCI classification. Compared with other state-of-the-art methods, the proposed framework achieves superior diagnosis accuracy, and hence could be promising for understanding pathological changes of brain connectome.

  View details for DOI 10.1038/s41598-017-06509-0

  View details for PubMedID 28747782

  View details for PubMedCentralID PMC5529469

• Optimized Motor Imagery Paradigm Based on Imagining Chinese Characters Writing Movement. IEEE transactions on neural systems and rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society Qiu, Z., Allison, B. Z., Jin, J., Zhang, Y., Wang, X., Li, W., Cichocki, A. 2017; 25 (7): 1009-1017

  Abstract

  motor imagery (MI) is a mental representation of motor behavior. The MI-based brain computer interfaces (BCIs) can provide communication for the physically impaired. The performance of MI-based BCI mainly depends on the subject's ability to self-modulate electroencephalogram signals. Proper training can help naive subjects learn to modulate brain activity proficiently. However, training subjects typically involve abstract motor tasks and are time-consuming.to improve the performance of naive subjects during motor imagery, a novel paradigm was presented that would guide naive subjects to modulate brain activity effectively. In this new paradigm, pictures of the left or right hand were used as cues for subjects to finish the motor imagery task. Fourteen healthy subjects (11 male, aged 22-25 years, and mean 23.6±1.16) participated in this study. The task was to imagine writing a Chinese character. Specifically, subjects could imagine hand movements corresponding to the sequence of writing strokes in the Chinese character. This paradigm was meant to find an effective and familiar action for most Chinese people, to provide them with a specific, extensively practiced task and help them modulate brain activity.results showed that the writing task paradigm yielded significantly better performance than the traditional arrow paradigm (p < 0.001). Questionnaire replies indicated that most subjects thought that the new paradigm was easier.the proposed new motor imagery paradigm could guide subjects to help them modulate brain activity effectively. Results showed that there were significant improvements using new paradigm, both in classification accuracy and usability.

  View details for DOI 10.1109/TNSRE.2017.2655542

  View details for PubMedID 28113345

• Drowsiness Detection by Bayesian-Copula Discriminant Classifier Based on EEG Signals During Daytime Short Nap. IEEE transactions on bio-medical engineering Qian, D., Wang, B., Qing, X., Zhang, T., Zhang, Y., Wang, X., Nakamura, M. 2017; 64 (4): 743-754

  Abstract

  Daytime short nap involves individual physiological states including alertness and drowsiness. In order to have a better understanding of the periodical rhymes of physiological states and then promote a good interpretability of alertness, the aim of this study is to detect drowsiness during daytime short nap.A method of Bayesian-copula discriminant classifier (BCDC) was introduced to detect individual drowsiness based on the physiological features extracted from electroencephalogram (EEG) signals. As an extension of traditional Bayesian decision theory, the BCDC method tries to construct the class-conditional probability density functions by exploiting the theory of copula and kernel density estimation.The proposed BCDC method was validated with experimental dataset and compared with other traditional methods for drowsiness detection. The obtained results showed that our method outperformed other methods in terms of three evaluation criteria.Our proposed method is effective to detect drowsiness with superior performance. Additionally, the BCDC method is relatively robust to different parameter settings on the group-level dataset.The proposed method is likely to be a useful tool to improve the correctness of the estimated class-conditional probability density functions. Since features are extracted from spontaneous EEG recordings, the results of this study can be further generalized to other experimental environment to detect vigilance level or driver drowsiness.

  View details for DOI 10.1109/TBME.2016.2574812

  View details for PubMedID 27254855

• Sparse Bayesian Learning for Obtaining Sparsity of EEG Frequency Bands Based Feature Vectors in Motor Imagery Classification. International journal of neural systems Zhang, Y., Wang, Y., Jin, J., Wang, X. 2017; 27 (2): 1650032

  Abstract

  Effective common spatial pattern (CSP) feature extraction for motor imagery (MI) electroencephalogram (EEG) recordings usually depends on the filter band selection to a large extent. Subband optimization has been suggested to enhance classification accuracy of MI. Accordingly, this study introduces a new method that implements sparse Bayesian learning of frequency bands (named SBLFB) from EEG for MI classification. CSP features are extracted on a set of signals that are generated by a filter bank with multiple overlapping subbands from raw EEG data. Sparse Bayesian learning is then exploited to implement selection of significant features with a linear discriminant criterion for classification. The effectiveness of SBLFB is demonstrated on the BCI Competition IV IIb dataset, in comparison with several other competing methods. Experimental results indicate that the SBLFB method is promising for development of an effective classifier to improve MI classification.

  View details for DOI 10.1142/S0129065716500325

  View details for PubMedID 27377661

• Sparse Bayesian multiway canonical correlation analysis for EEG pattern recognition NEUROCOMPUTING Zhang, Y., Zhou, G., Jin, J., Zhang, Y., Wang, X., Cichocki, A. 2017; 225: 103-110

  View details for DOI 10.1016/j.neucom.2016.11.008

  View details for Web of Science ID 000392164400010

• Robust frequency recognition for SSVEP-based BCI with temporally local multivariate synchronization index. Cognitive neurodynamics Zhang, Y., Guo, D., Xu, P., Zhang, Y., Yao, D. 2016; 10 (6): 505-511

  Abstract

  Multivariate synchronization index (MSI) has been proved to be an efficient method for frequency recognition in SSVEP-BCI systems. It measures the correlation according to the entropy of the normalized eigenvalues of the covariance matrix of multichannel signals. In the MSI method, the estimation of covariance matrix omits the temporally local structure of samples. In this study, a new spatio-temporal method, termed temporally local MSI (TMSI), was presented. This new method explicitly exploits temporally local information in modelling the covariance matrix. In order to evaluate the performance of the TMSI, we performs a comparison between the two methods on the real SSVEP datasets from eleven subjects. The results show that the TMSI outperforms the standard MSI. TMSI benefits from exploiting the temporally local structure of EEG signals, and could be a potential method for robust performance of SSVEP-based BCI.

  View details for DOI 10.1007/s11571-016-9398-9

  View details for PubMedID 27891199

  View details for PubMedCentralID PMC5106453

• Group Component Analysis for Multiblock Data: Common and Individual Feature Extraction. IEEE transactions on neural networks and learning systems Zhou, G., Cichocki, A., Zhang, Y., Mandic, D. P. 2016; 27 (11): 2426-2439

  Abstract

  Real-world data are often acquired as a collection of matrices rather than as a single matrix. Such multiblock data are naturally linked and typically share some common features while at the same time exhibiting their own individual features, reflecting the underlying data generation mechanisms. To exploit the linked nature of data, we propose a new framework for common and individual feature extraction (CIFE) which identifies and separates the common and individual features from the multiblock data. Two efficient algorithms termed common orthogonal basis extraction (COBE) are proposed to extract common basis is shared by all data, independent on whether the number of common components is known beforehand. Feature extraction is then performed on the common and individual subspaces separately, by incorporating dimensionality reduction and blind source separation techniques. Comprehensive experimental results on both the synthetic and real-world data demonstrate significant advantages of the proposed CIFE method in comparison with the state-of-the-art.

  View details for DOI 10.1109/TNNLS.2015.2487364

  View details for PubMedID 26529787

• Sparse Bayesian Classification of EEG for Brain-Computer Interface. IEEE transactions on neural networks and learning systems Zhang, Y., Zhou, G., Jin, J., Zhao, Q., Wang, X., Cichocki, A. 2016; 27 (11): 2256-2267

  Abstract

  Regularization has been one of the most popular approaches to prevent overfitting in electroencephalogram (EEG) classification of brain-computer interfaces (BCIs). The effectiveness of regularization is often highly dependent on the selection of regularization parameters that are typically determined by cross-validation (CV). However, the CV imposes two main limitations on BCIs: 1) a large amount of training data is required from the user and 2) it takes a relatively long time to calibrate the classifier. These limitations substantially deteriorate the system's practicability and may cause a user to be reluctant to use BCIs. In this paper, we introduce a sparse Bayesian method by exploiting Laplace priors, namely, SBLaplace, for EEG classification. A sparse discriminant vector is learned with a Laplace prior in a hierarchical fashion under a Bayesian evidence framework. All required model parameters are automatically estimated from training data without the need of CV. Extensive comparisons are carried out between the SBLaplace algorithm and several other competing methods based on two EEG data sets. The experimental results demonstrate that the SBLaplace algorithm achieves better overall performance than the competing algorithms for EEG classification.

  View details for DOI 10.1109/TNNLS.2015.2476656

  View details for PubMedID 26415189

• Improved SFFS method for channel selection in motor imagery based BCI NEUROCOMPUTING Qiu, Z., Jin, J., Lam, H., Zhang, Y., Wang, X., Cichocki, A. 2016; 207: 519-527

  View details for DOI 10.1016/j.neucom.2016.05.035

  View details for Web of Science ID 000382794500048

• Fast nonnegative tensor factorization based on accelerated proximal gradient and low-rank approximation Zhang, Y., Zhou, G., Zhao, Q., Cichocki, A., Wang, X. ELSEVIER. 2016: 148-154

  View details for DOI 10.1016/j.neucom.2015.08.122

  View details for Web of Science ID 000377230300019

• An exploration of spatial auditory BCI paradigms with different sounds: music notes versus beeps. Cognitive neurodynamics Huang, M., Daly, I., Jin, J., Zhang, Y., Wang, X., Cichocki, A. 2016; 10 (3): 201-9

  Abstract

  Visual brain-computer interfaces (BCIs) are not suitable for people who cannot reliably maintain their eye gaze. Considering that this group usually maintains audition, an auditory based BCI may be a good choice for them. In this paper, we explore two auditory patterns: (1) a pattern utilizing symmetrical spatial cues with multiple frequency beeps [called the high low medium (HLM) pattern], and (2) a pattern utilizing non-symmetrical spatial cues with six tones derived from the diatonic scale [called the diatonic scale (DS) pattern]. These two patterns are compared to each other in terms of accuracy to determine which auditory pattern is better. The HLM pattern uses three different frequency beeps and has a symmetrical spatial distribution. The DS pattern uses six spoken stimuli, which are six notes solmizated as "do", "re", "mi", "fa", "sol" and "la", and derived from the diatonic scale. These six sounds are distributed to six, spatially distributed, speakers. Thus, we compare a BCI paradigm using beeps with another BCI paradigm using tones on the diatonic scale, when the stimuli are spatially distributed. Although no significant differences are found between the ERPs, the HLM pattern performs better than the DS pattern: the online accuracy achieved with the HLM pattern is significantly higher than that achieved with the DS pattern (p = 0.0028).

  View details for DOI 10.1007/s11571-016-9377-1

  View details for PubMedID 27275376

  View details for PubMedCentralID PMC4870409

• Discriminative Feature Extraction via Multivariate Linear Regression for SSVEP-Based BCI. IEEE transactions on neural systems and rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society Wang, H., Zhang, Y., Waytowich, N. R., Krusienski, D. J., Zhou, G., Jin, J., Wang, X., Cichocki, A. 2016; 24 (5): 532-41

  Abstract

  Many of the most widely accepted methods for reliable detection of steady-state visual evoked potentials (SSVEPs) in the electroencephalogram (EEG) utilize canonical correlation analysis (CCA). CCA uses pure sine and cosine reference templates with frequencies corresponding to the visual stimulation frequencies. These generic reference templates may not optimally reflect the natural SSVEP features obscured by the background EEG. This paper introduces a new approach that utilizes spatio-temporal feature extraction with multivariate linear regression (MLR) to learn discriminative SSVEP features for improving the detection accuracy. MLR is implemented on dimensionality-reduced EEG training data and a constructed label matrix to find optimally discriminative subspaces. Experimental results show that the proposed MLR method significantly outperforms CCA as well as several other competing methods for SSVEP detection, especially for time windows shorter than 1 second. This demonstrates that the MLR method is a promising new approach for achieving improved real-time performance of SSVEP-BCIs.

  View details for DOI 10.1109/TNSRE.2016.2519350

  View details for PubMedID 26812728

• Linked Component Analysis From Matrices to High-Order Tensors: Applications to Biomedical Data PROCEEDINGS OF THE IEEE Zhou, G., Zhao, Q., Zhang, Y., Adali, T., Xie, S., Cichocki, A. 2016; 104 (2): 310-331

  View details for DOI 10.1109/JPROC.2015.2474704

  View details for Web of Science ID 000368980100007

• Exploring Combinations of Different Color and Facial Expression Stimuli for Gaze-Independent BCIs. Frontiers in computational neuroscience Chen, L., Jin, J., Daly, I., Zhang, Y., Wang, X., Cichocki, A. 2016; 10: 5

  Abstract

  Some studies have proven that a conventional visual brain computer interface (BCI) based on overt attention cannot be used effectively when eye movement control is not possible. To solve this problem, a novel visual-based BCI system based on covert attention and feature attention has been proposed and was called the gaze-independent BCI. Color and shape difference between stimuli and backgrounds have generally been used in examples of gaze-independent BCIs. Recently, a new paradigm based on facial expression changes has been presented, and obtained high performance. However, some facial expressions were so similar that users couldn't tell them apart, especially when they were presented at the same position in a rapid serial visual presentation (RSVP) paradigm. Consequently, the performance of the BCI is reduced.In this paper, we combined facial expressions and colors to optimize the stimuli presentation in the gaze-independent BCI. This optimized paradigm was called the colored dummy face pattern. It is suggested that different colors and facial expressions could help users to locate the target and evoke larger event-related potentials (ERPs). In order to evaluate the performance of this new paradigm, two other paradigms were presented, called the gray dummy face pattern and the colored ball pattern.The key point that determined the value of the colored dummy faces stimuli in BCI systems was whether the dummy face stimuli could obtain higher performance than gray faces or colored balls stimuli. Ten healthy participants (seven male, aged 21-26 years, mean 24.5 ± 1.25) participated in our experiment. Online and offline results of four different paradigms were obtained and comparatively analyzed.The results showed that the colored dummy face pattern could evoke higher P300 and N400 ERP amplitudes, compared with the gray dummy face pattern and the colored ball pattern. Online results showed that the colored dummy face pattern had a significant advantage in terms of classification accuracy (p < 0.05) and information transfer rate (p < 0.05) compared to the other two patterns.The stimuli used in the colored dummy face paradigm combined color and facial expressions. This had a significant advantage in terms of the evoked P300 and N400 amplitudes and resulted in high classification accuracies and information transfer rates. It was compared with colored ball and gray dummy face stimuli.

  View details for DOI 10.3389/fncom.2016.00005

  View details for PubMedID 26858634

  View details for PubMedCentralID PMC4731496

• Optimizing spatial patterns with sparse filter bands for motor-imagery based brain-computer interface. Journal of neuroscience methods Zhang, Y., Zhou, G., Jin, J., Wang, X., Cichocki, A. 2015; 255: 85-91

  Abstract

  Common spatial pattern (CSP) has been most popularly applied to motor-imagery (MI) feature extraction for classification in brain-computer interface (BCI) application. Successful application of CSP depends on the filter band selection to a large degree. However, the most proper band is typically subject-specific and can hardly be determined manually.This study proposes a sparse filter band common spatial pattern (SFBCSP) for optimizing the spatial patterns. SFBCSP estimates CSP features on multiple signals that are filtered from raw EEG data at a set of overlapping bands. The filter bands that result in significant CSP features are then selected in a supervised way by exploiting sparse regression. A support vector machine (SVM) is implemented on the selected features for MI classification.Two public EEG datasets (BCI Competition III dataset IVa and BCI Competition IV IIb) are used to validate the proposed SFBCSP method. Experimental results demonstrate that SFBCSP help improve the classification performance of MI.The optimized spatial patterns by SFBCSP give overall better MI classification accuracy in comparison with several competing methods.The proposed SFBCSP is a potential method for improving the performance of MI-based BCI.

  View details for DOI 10.1016/j.jneumeth.2015.08.004

  View details for PubMedID 26277421

• A P300 brain-computer interface based on a modification of the mismatch negativity paradigm. International journal of neural systems Jin, J., Sellers, E. W., Zhou, S., Zhang, Y., Wang, X., Cichocki, A. 2015; 25 (3): 1550011

  Abstract

  The P300-based brain-computer interface (BCI) is an extension of the oddball paradigm, and can facilitate communication for people with severe neuromuscular disorders. It has been shown that, in addition to the P300, other event-related potential (ERP) components have been shown to contribute to successful operation of the P300 BCI. Incorporating these components into the classification algorithm can improve the classification accuracy and information transfer rate (ITR). In this paper, a single character presentation paradigm was compared to a presentation paradigm that is based on the visual mismatch negativity. The mismatch negativity paradigm showed significantly higher classification accuracy and ITRs than a single character presentation paradigm. In addition, the mismatch paradigm elicited larger N200 and N400 components than the single character paradigm. The components elicited by the presentation method were consistent with what would be expected from a mismatch paradigm and a typical P300 was also observed. The results show that increasing the signal-to-noise ratio by increasing the amplitude of ERP components can significantly improve BCI speed and accuracy. The mismatch presentation paradigm may be considered a viable option to the traditional P300 BCI paradigm.

  View details for DOI 10.1142/S0129065715500112

  View details for PubMedID 25804352

• A new hybrid BCI paradigm based on P300 and SSVEP. Journal of neuroscience methods Wang, M., Daly, I., Allison, B. Z., Jin, J., Zhang, Y., Chen, L., Wang, X. 2015; 244: 16-25

  Abstract

  P300 and steady-state visual evoked potential (SSVEP) approaches have been widely used for brain-computer interface (BCI) systems. However, neither of these approaches can work for all subjects. Some groups have reported that a hybrid BCI that combines two or more approaches might provide BCI functionality to more users. Hybrid P300/SSVEP BCIs have only recently been developed and validated, and very few avenues to improve performance have been explored.The present study compares an established hybrid P300/SSVEP BCIs paradigm to a new paradigm in which shape changing, instead of color changing, is adopted for P300 evocation to decrease the degradation on SSVEP strength.The result shows that the new hybrid paradigm presented in this paper yields much better performance than the normal hybrid paradigm.A performance increase of nearly 20% in SSVEP classification is achieved using the new hybrid paradigm in comparison with the normal hybrid paradigm. All the paradigms except the normal hybrid paradigm used in this paper obtain 100% accuracy in P300 classification.The new hybrid P300/SSVEP BCIs paradigm in which shape changing, instead of color changing, could obtain as high classification accuracy of SSVEP as the traditional SSVEP paradigm and could obtain as high classification accuracy of P300 as the traditional P300 paradigm. P300 did not interfere with the SSVEP response using the new hybrid paradigm presented in this paper, which was superior to the normal hybrid P300/SSVEP paradigm.

  View details for DOI 10.1016/j.jneumeth.2014.06.003

  View details for PubMedID 24997343

• SSVEP recognition using common feature analysis in brain-computer interface. Journal of neuroscience methods Zhang, Y., Zhou, G., Jin, J., Wang, X., Cichocki, A. 2015; 244: 8-15

  Abstract

  Canonical correlation analysis (CCA) has been successfully applied to steady-state visual evoked potential (SSVEP) recognition for brain-computer interface (BCI) application. Although the CCA method outperforms the traditional power spectral density analysis through multi-channel detection, it requires additionally pre-constructed reference signals of sine-cosine waves. It is likely to encounter overfitting in using a short time window since the reference signals include no features from training data.We consider that a group of electroencephalogram (EEG) data trials recorded at a certain stimulus frequency on a same subject should share some common features that may bear the real SSVEP characteristics. This study therefore proposes a common feature analysis (CFA)-based method to exploit the latent common features as natural reference signals in using correlation analysis for SSVEP recognition.Good performance of the CFA method for SSVEP recognition is validated with EEG data recorded from ten healthy subjects, in contrast to CCA and a multiway extension of CCA (MCCA).Experimental results indicate that the CFA method significantly outperformed the CCA and the MCCA methods for SSVEP recognition in using a short time window (i.e., less than 1s).The superiority of the proposed CFA method suggests it is promising for the development of a real-time SSVEP-based BCI.

  View details for DOI 10.1016/j.jneumeth.2014.03.012

  View details for PubMedID 24727656

• A novel EOG/EEG hybrid human-machine interface adopting eye movements and ERPs: application to robot control. IEEE transactions on bio-medical engineering Ma, J., Zhang, Y., Cichocki, A., Matsuno, F. 2015; 62 (3): 876-89

  Abstract

  This study presents a novel human-machine interface (HMI) based on both electrooculography (EOG) and electroencephalography (EEG). This hybrid interface works in two modes: an EOG mode recognizes eye movements such as blinks, and an EEG mode detects event related potentials (ERPs) like P300. While both eye movements and ERPs have been separately used for implementing assistive interfaces, which help patients with motor disabilities in performing daily tasks, the proposed hybrid interface integrates them together. In this way, both the eye movements and ERPs complement each other. Therefore, it can provide a better efficiency and a wider scope of application. In this study, we design a threshold algorithm that can recognize four kinds of eye movements including blink, wink, gaze, and frown. In addition, an oddball paradigm with stimuli of inverted faces is used to evoke multiple ERP components including P300, N170, and VPP. To verify the effectiveness of the proposed system, two different online experiments are carried out. One is to control a multifunctional humanoid robot, and the other is to control four mobile robots. In both experiments, the subjects can complete tasks effectively by using the proposed interface, whereas the best completion time is relatively short and very close to the one operated by hand.

  View details for DOI 10.1109/TBME.2014.2369483

  View details for PubMedID 25398172

• A survey of the dummy face and human face stimuli used in BCI paradigm. Journal of neuroscience methods Chen, L., Jin, J., Zhang, Y., Wang, X., Cichocki, A. 2015; 239: 18-27

  Abstract

  It was proved that the human face stimulus were superior to the flash only stimulus in BCI system. However, human face stimulus may lead to copyright infringement problems and was hard to be edited according to the requirement of the BCI study. Recently, it was reported that facial expression changes could be done by changing a curve in a dummy face which could obtain good performance when it was applied to visual-based P300 BCI systems.In this paper, four different paradigms were presented, which were called dummy face pattern, human face pattern, inverted dummy face pattern and inverted human face pattern, to evaluate the performance of the dummy faces stimuli compared with the human faces stimuli.The key point that determined the value of dummy faces in BCI systems were whether dummy faces stimuli could obtain as good performance as human faces stimuli. Online and offline results of four different paradigms would have been obtained and comparatively analyzed.Online and offline results showed that there was no significant difference among dummy faces and human faces in ERPs, classification accuracy and information transfer rate when they were applied in BCI systems.Dummy faces stimuli could evoke large ERPs and obtain as high classification accuracy and information transfer rate as the human faces stimuli. Since dummy faces were easy to be edited and had no copyright infringement problems, it would be a good choice for optimizing the stimuli of BCI systems.

  View details for DOI 10.1016/j.jneumeth.2014.10.002

  View details for PubMedID 25314905

• An ERP-based BCI using an oddball paradigm with different faces and reduced errors in critical functions. International journal of neural systems Jin, J., Allison, B. Z., Zhang, Y., Wang, X., Cichocki, A. 2014; 24 (8): 1450027

  Abstract

  Recent research has shown that a new face paradigm is superior to the conventional "flash only" approach that has dominated P300 brain-computer interfaces (BCIs) for over 20 years. However, these face paradigms did not study the repetition effects and the stability of evoked event related potentials (ERPs), which would decrease the performance of P300 BCI. In this paper, we explored whether a new "multi-faces (MF)" approach would yield more distinct ERPs than the conventional "single face (SF)" approach. To decrease the repetition effects and evoke large ERPs, we introduced a new stimulus approach called the "MF" approach, which shows different familiar faces randomly. Fifteen subjects participated in runs using this new approach and an established "SF" approach. The result showed that the MF pattern enlarged the N200 and N400 components, evoked stable P300 and N400, and yielded better BCI performance than the SF pattern. The MF pattern can evoke larger N200 and N400 components and more stable P300 and N400, which increase the classification accuracy compared to the face pattern.

  View details for DOI 10.1142/S0129065714500270

  View details for PubMedID 25182191

• An efficient frequency recognition method based on likelihood ratio test for SSVEP-based BCI. Computational and mathematical methods in medicine Zhang, Y., Dong, L., Zhang, R., Yao, D., Zhang, Y., Xu, P. 2014; 2014: 908719

  Abstract

  An efficient frequency recognition method is very important for SSVEP-based BCI systems to improve the information transfer rate (ITR). To address this aspect, for the first time, likelihood ratio test (LRT) was utilized to propose a novel multichannel frequency recognition method for SSVEP data. The essence of this new method is to calculate the association between multichannel EEG signals and the reference signals which were constructed according to the stimulus frequency with LRT. For the simulation and real SSVEP data, the proposed method yielded higher recognition accuracy with shorter time window length and was more robust against noise in comparison with the popular canonical correlation analysis- (CCA-) based method and the least absolute shrinkage and selection operator- (LASSO-) based method. The recognition accuracy and information transfer rate (ITR) obtained by the proposed method was higher than those of the CCA-based method and LASSO-based method. The superior results indicate that the LRT method is a promising candidate for reliable frequency recognition in future SSVEP-BCI.

  View details for DOI 10.1155/2014/908719

  View details for PubMedID 25250058

  View details for PubMedCentralID PMC4163431

• An optimized ERP brain-computer interface based on facial expression changes. Journal of neural engineering Jin, J., Daly, I., Zhang, Y., Wang, X., Cichocki, A. 2014; 11 (3): 036004

  Abstract

  Interferences from spatially adjacent non-target stimuli are known to evoke event-related potentials (ERPs) during non-target flashes and, therefore, lead to false positives. This phenomenon was commonly seen in visual attention-based brain-computer interfaces (BCIs) using conspicuous stimuli and is known to adversely affect the performance of BCI systems. Although users try to focus on the target stimulus, they cannot help but be affected by conspicuous changes of the stimuli (such as flashes or presenting images) which were adjacent to the target stimulus. Furthermore, subjects have reported that conspicuous stimuli made them tired and annoyed. In view of this, the aim of this study was to reduce adjacent interference, annoyance and fatigue using a new stimulus presentation pattern based upon facial expression changes. Our goal was not to design a new pattern which could evoke larger ERPs than the face pattern, but to design a new pattern which could reduce adjacent interference, annoyance and fatigue, and evoke ERPs as good as those observed during the face pattern.Positive facial expressions could be changed to negative facial expressions by minor changes to the original facial image. Although the changes are minor, the contrast is big enough to evoke strong ERPs. In this paper, a facial expression change pattern between positive and negative facial expressions was used to attempt to minimize interference effects. This was compared against two different conditions, a shuffled pattern containing the same shapes and colours as the facial expression change pattern, but without the semantic content associated with a change in expression, and a face versus no face pattern. Comparisons were made in terms of classification accuracy and information transfer rate as well as user supplied subjective measures.The results showed that interferences from adjacent stimuli, annoyance and the fatigue experienced by the subjects could be reduced significantly (p < 0.05) by using the facial expression change patterns in comparison with the face pattern. The offline results show that the classification accuracy of the facial expression change pattern was significantly better than that of the shuffled pattern (p < 0.05) and the face pattern (p < 0.05).The facial expression change pattern presented in this paper reduced interference from adjacent stimuli and decreased the fatigue and annoyance experienced by BCI users significantly (p < 0.05) compared to the face pattern.

  View details for DOI 10.1088/1741-2560/11/3/036004

  View details for PubMedID 24743165

• Frequency recognition in SSVEP-based BCI using multiset canonical correlation analysis. International journal of neural systems Zhang, Y., Zhou, G., Jin, J., Wang, X., Cichocki, A. 2014; 24 (4): 1450013

  Abstract

  Canonical correlation analysis (CCA) has been one of the most popular methods for frequency recognition in steady-state visual evoked potential (SSVEP)-based brain-computer interfaces (BCIs). Despite its efficiency, a potential problem is that using pre-constructed sine-cosine waves as the required reference signals in the CCA method often does not result in the optimal recognition accuracy due to their lack of features from the real electro-encephalo-gram (EEG) data. To address this problem, this study proposes a novel method based on multiset canonical correlation analysis (MsetCCA) to optimize the reference signals used in the CCA method for SSVEP frequency recognition. The MsetCCA method learns multiple linear transforms that implement joint spatial filtering to maximize the overall correlation among canonical variates, and hence extracts SSVEP common features from multiple sets of EEG data recorded at the same stimulus frequency. The optimized reference signals are formed by combination of the common features and completely based on training data. Experimental study with EEG data from 10 healthy subjects demonstrates that the MsetCCA method improves the recognition accuracy of SSVEP frequency in comparison with the CCA method and other two competing methods (multiway CCA (MwayCCA) and phase constrained CCA (PCCA)), especially for a small number of channels and a short time window length. The superiority indicates that the proposed MsetCCA method is a new promising candidate for frequency recognition in SSVEP-based BCIs.

  View details for DOI 10.1142/S0129065714500130

  View details for PubMedID 24694168

• Aggregation of sparse linear discriminant analyses for event-related potential classification in brain-computer interface. International journal of neural systems Zhang, Y., Zhou, G., Jin, J., Zhao, Q., Wang, X., Cichocki, A. 2014; 24 (1): 1450003

  Abstract

  Two main issues for event-related potential (ERP) classification in brain-computer interface (BCI) application are curse-of-dimensionality and bias-variance tradeoff, which may deteriorate classification performance, especially with insufficient training samples resulted from limited calibration time. This study introduces an aggregation of sparse linear discriminant analyses (ASLDA) to overcome these problems. In the ASLDA, multiple sparse discriminant vectors are learned from differently l1-regularized least-squares regressions by exploiting the equivalence between LDA and least-squares regression, and are subsequently aggregated to form an ensemble classifier, which could not only implement automatic feature selection for dimensionality reduction to alleviate curse-of-dimensionality, but also decrease the variance to improve generalization capacity for new test samples. Extensive investigation and comparison are carried out among the ASLDA, the ordinary LDA and other competing ERP classification algorithms, based on different three ERP datasets. Experimental results indicate that the ASLDA yields better overall performance for single-trial ERP classification when insufficient training samples are available. This suggests the proposed ASLDA is promising for ERP classification in small sample size scenario to improve the practicability of BCI.

  View details for DOI 10.1142/S0129065714500038

  View details for PubMedID 24344691

• A four-choice hybrid P300/SSVEP BCI for improved accuracy BRAIN-COMPUTER INTERFACES Allison, B. Z., Jin, J., Zhang, Y., Wang, X. 2014; 1 (1): 17-26

  View details for DOI 10.1080/2326263X.2013.869003

  View details for Web of Science ID 000218513300004

• L1-regularized Multiway canonical correlation analysis for SSVEP-based BCI. IEEE transactions on neural systems and rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society Zhang, Y., Zhou, G., Jin, J., Wang, M., Wang, X., Cichocki, A. 2013; 21 (6): 887-96

  Abstract

  Canonical correlation analysis (CCA) between recorded electroencephalogram (EEG) and designed reference signals of sine-cosine waves usually works well for steady-state visual evoked potential (SSVEP) recognition in brain-computer interface (BCI) application. However, using the reference signals of sine- cosine waves without subject-specific and inter-trial information can hardly give the optimal recognition accuracy, due to possible overfitting, especially within a short time window length. This paper introduces an L1-regularized multiway canonical correlation analysis (L1-MCCA) for reference signal optimization to improve the SSVEP recognition performance further. A multiway extension of the CCA, called MCCA, is first presented, in which collaborative CCAs are exploited to optimize the reference signals in correlation analysis for SSVEP recognition alternatingly from the channel-way and trial-way arrays of constructed EEG tensor. L1-regularization is subsequently imposed on the trial-way array optimization in the MCCA, and hence results in the more powerful L1-MCCA with function of effective trial selection. Both the proposed MCCA and L1-MCCA methods are validated for SSVEP recognition with EEG data from 10 healthy subjects, and compared to the ordinary CCA without reference signal optimization. Experimental results show that the MCCA significantly outperforms the CCA for SSVEP recognition. The L1-MCCA further improves the recognition accuracy which is significantly higher than that of the MCCA.

  View details for DOI 10.1109/TNSRE.2013.2279680

  View details for PubMedID 24122565

• Spatial-temporal discriminant analysis for ERP-based brain-computer interface. IEEE transactions on neural systems and rehabilitation engineering : a publication of the IEEE Engineering in Medicine and Biology Society Zhang, Y., Zhou, G., Zhao, Q., Jin, J., Wang, X., Cichocki, A. 2013; 21 (2): 233-43

  Abstract

  Linear discriminant analysis (LDA) has been widely adopted to classify event-related potential (ERP) in brain-computer interface (BCI). Good classification performance of the ERP-based BCI usually requires sufficient data recordings for effective training of the LDA classifier, and hence a long system calibration time which however may depress the system practicability and cause the users resistance to the BCI system. In this study, we introduce a spatial-temporal discriminant analysis (STDA) to ERP classification. As a multiway extension of the LDA, the STDA method tries to maximize the discriminant information between target and nontarget classes through finding two projection matrices from spatial and temporal dimensions collaboratively, which reduces effectively the feature dimensionality in the discriminant analysis, and hence decreases significantly the number of required training samples. The proposed STDA method was validated with dataset II of the BCI Competition III and dataset recorded from our own experiments, and compared to the state-of-the-art algorithms for ERP classification. Online experiments were additionally implemented for the validation. The superior classification performance in using few training samples shows that the STDA is effective to reduce the system calibration time and improve the classification accuracy, thereby enhancing the practicability of ERP-based BCI.

  View details for DOI 10.1109/TNSRE.2013.2243471

  View details for PubMedID 23476005

• Whether generic model works for rapid ERP-based BCI calibration. Journal of neuroscience methods Jin, J., Sellers, E. W., Zhang, Y., Daly, I., Wang, X., Cichocki, A. 2013; 212 (1): 94-9

  Abstract

  Event-related potential (ERP)-based brain-computer interfacing (BCI) is an effective method of basic communication. However, collecting calibration data, and classifier training, detracts from the amount of time allocated for online communication. Decreasing calibration time can reduce preparation time thereby allowing for additional online use, potentially lower fatigue, and improved performance. Previous studies, using generic online training models which avoid offline calibration, afford more time for online spelling. Such studies have not examined the direct effects of the model on individual performance, and the training sequence exceeded the time reported here. The first goal of this work is to survey whether one generic model works for all subjects and the second goal is to show the performance of a generic model using an online training strategy when participants could use the generic model. The generic model was derived from 10 participant's data. An additional 11 participants were recruited for the current study. Seven of the participants were able to use the generic model during online training. Moreover, the generic model performed as well as models obtained from participant specific offline data with a mean training time of less than 2 min. However, four of the participants could not use this generic model, which shows that one generic mode is not generic for all subjects. More research on ERPs of subjects with different characteristics should be done, which would be helpful to build generic models for subject groups. This result shows a potential valuable direction for improving the BCI system.

  View details for DOI 10.1016/j.jneumeth.2012.09.020

  View details for PubMedID 23032116

  View details for PubMedCentralID PMC3658461

• The changing face of P300 BCIs: a comparison of stimulus changes in a P300 BCI involving faces, emotion, and movement. PloS one Jin, J., Allison, B. Z., Kaufmann, T., Kübler, A., Zhang, Y., Wang, X., Cichocki, A. 2012; 7 (11): e49688

  Abstract

  One of the most common types of brain-computer interfaces (BCIs) is called a P300 BCI, since it relies on the P300 and other event-related potentials (ERPs). In the canonical P300 BCI approach, items on a monitor flash briefly to elicit the necessary ERPs. Very recent work has shown that this approach may yield lower performance than alternate paradigms in which the items do not flash but instead change in other ways, such as moving, changing colour or changing to characters overlaid with faces.The present study sought to extend this research direction by parametrically comparing different ways to change items in a P300 BCI. Healthy subjects used a P300 BCI across six different conditions. Three conditions were similar to our prior work, providing the first direct comparison of characters flashing, moving, and changing to faces. Three new conditions also explored facial motion and emotional expression. The six conditions were compared across objective measures such as classification accuracy and bit rate as well as subjective measures such as perceived difficulty. In line with recent studies, our results indicated that the character flash condition resulted in the lowest accuracy and bit rate. All four face conditions (mean accuracy >91%) yielded significantly better performance than the flash condition (mean accuracy = 75%).Objective results reaffirmed that the face paradigm is superior to the canonical flash approach that has dominated P300 BCIs for over 20 years. The subjective reports indicated that the conditions that yielded better performance were not considered especially burdensome. Therefore, although further work is needed to identify which face paradigm is best, it is clear that the canonical flash approach should be replaced with a face paradigm when aiming at increasing bit rate. However, the face paradigm has to be further explored with practical applications particularly with locked-in patients.

  View details for DOI 10.1371/journal.pone.0049688

  View details for PubMedID 23189154

  View details for PubMedCentralID PMC3506655

• Assessment of human operator functional state using a novel differential evolution optimization based adaptive fuzzy model BIOMEDICAL SIGNAL PROCESSING AND CONTROL Wang, R., Zhang, J., Zhang, Y., Wang, X. 2012; 7 (5): 490-498

  View details for DOI 10.1016/j.bspc.2011.09.004

  View details for Web of Science ID 000307135800009

• A novel BCI based on ERP components sensitive to configural processing of human faces. Journal of neural engineering Zhang, Y., Zhao, Q., Jin, J., Wang, X., Cichocki, A. 2012; 9 (2): 026018

  Abstract

  This study introduces a novel brain-computer interface (BCI) based on an oddball paradigm using stimuli of facial images with loss of configural face information (e.g., inversion of face). To the best of our knowledge, till now the configural processing of human faces has not been applied to BCI but widely studied in cognitive neuroscience research. Our experiments confirm that the face-sensitive event-related potential (ERP) components N170 and vertex positive potential (VPP) have reflected early structural encoding of faces and can be modulated by the configural processing of faces. With the proposed novel paradigm, we investigate the effects of ERP components N170, VPP and P300 on target detection for BCI. An eight-class BCI platform is developed to analyze ERPs and evaluate the target detection performance using linear discriminant analysis without complicated feature extraction processing. The online classification accuracy of 88.7% and information transfer rate of 38.7 bits min(-1) using stimuli of inverted faces with only single trial suggest that the proposed paradigm based on the configural processing of faces is very promising for visual stimuli-driven BCI applications.

  View details for DOI 10.1088/1741-2560/9/2/026018

  View details for PubMedID 22414683

• LASSO based stimulus frequency recognition model for SSVEP BCIs BIOMEDICAL SIGNAL PROCESSING AND CONTROL Zhang, Y., Jin, J., Qing, X., Wang, B., Wang, X. 2012; 7 (2): 104-111

  View details for DOI 10.1016/j.bspc.2011.02.002

  View details for Web of Science ID 000301759300002