Coil sketching for computationally efficient MR iterative reconstruction.
Magnetic resonance in medicine
Parallel imaging and compressed sensing reconstructions of large MRI datasets often have a prohibitive computational cost that bottlenecks clinical deployment, especially for three-dimensional (3D) non-Cartesian acquisitions. One common approach is to reduce the number of coil channels actively used during reconstruction as in coil compression. While effective for Cartesian imaging, coil compression inherently loses signal energy, producing shading artifacts that compromise image quality for 3D non-Cartesian imaging. We propose coil sketching, a general and versatile method for computationally-efficient iterative MR image reconstruction.We based our method on randomized sketching algorithms, a type of large-scale optimization algorithms well established in the fields of machine learning and big data analysis. We adapt the sketching theory to the MRI reconstruction problem via a structured sketching matrix that, similar to coil compression, considers high-energy virtual coils obtained from principal component analysis. But, unlike coil compression, it also considers random linear combinations of the remaining low-energy coils, effectively leveraging information from all coils.First, we performed ablation experiments to validate the sketching matrix design on both Cartesian and non-Cartesian datasets. The resulting design yielded both improved computatioanal efficiency and preserved signal-to-noise ratio (SNR) as measured by the inverse g-factor. Then, we verified the efficacy of our approach on high-dimensional non-Cartesian 3D cones datasets, where coil sketching yielded up to three-fold faster reconstructions with equivalent image quality.Coil sketching is a general and versatile reconstruction framework for computationally fast and memory-efficient reconstruction.
View details for DOI 10.1002/mrm.29883
View details for PubMedID 37848365
Noise2Recon: Enabling SNR-robust MRI reconstruction with semi-supervised and self-supervised learning.
Magnetic resonance in medicine
PURPOSE: To develop a method for building MRI reconstruction neural networks robust to changes in signal-to-noise ratio (SNR) and trainable with a limited number of fully sampled scans.METHODS: We propose Noise2Recon, a consistency training method for SNR-robust accelerated MRI reconstruction that can use both fully sampled (labeled) and undersampled (unlabeled) scans. Noise2Recon uses unlabeled data by enforcing consistency between model reconstructions of undersampled scans and their noise-augmented counterparts. Noise2Recon was compared to compressed sensing and both supervised and self-supervised deep learning baselines. Experiments were conducted using retrospectively accelerated data from the mridata three-dimensional fast-spin-echo knee and two-dimensional fastMRI brain datasets. All methods were evaluated in label-limited settings and among out-of-distribution (OOD) shifts, including changes in SNR, acceleration factors, and datasets. An extensive ablation study was conducted to characterize the sensitivity of Noise2Recon to hyperparameter choices.RESULTS: In label-limited settings, Noise2Recon achieved better structural similarity, peak signal-to-noise ratio, and normalized-RMS error than all baselines and matched performance of supervised models, which were trained with 14 * $$ 14\times $$ more fully sampled scans. Noise2Recon outperformed all baselines, including state-of-the-art fine-tuning and augmentation techniques, among low-SNR scans and when generalizing to OOD acceleration factors. Augmentation extent and loss weighting hyperparameters had negligible impact on Noise2Recon compared to supervised methods, which may indicate increased training stability.CONCLUSION: Noise2Recon is a label-efficient reconstruction method that is robust to distribution shifts, such as changes in SNR, acceleration factors, and others, with limited or no fully sampled training data.
View details for DOI 10.1002/mrm.29759
View details for PubMedID 37427449
Unraveling Attention via Convex Duality: Analysis and Interpretations of Vision Transformers
JMLR-JOURNAL MACHINE LEARNING RESEARCH. 2022: 19050-19088
View details for Web of Science ID 000900130200004
Musical mirror-symmetrical movement tasks: comparison of rhythm versus melody-playing.
Bimanual mirror-symmetrical movement (MSM) is relatively easy to control movement. Different MSM tasks may have different activations and interhemispheric interactions. The purpose of this study is to compare anatomo-physiological features such as hemispheric activations and dominance of two different MSMs, namely melody-playing and rhythm. We examined functional MRI (fMRI) recordings in a group of fifteen right-handed pianists performing two separate tasks: bimanual rhythm and bimanual melody-playing on two different keyboards with standard key order for right hand and reversed for left hand, which allows homolog fingers' movements. Activations and laterality indices on fMRI were examined. The results show that significant cerebellar activations (especially in anterior cerebellum) in both groups. Significant primary sensorimotor cortical activations are observed in the melody-playing group. While there are also bilaterally symmetric activations, and laterality indices suggest overall lateralization towards the left hemisphere in both groups. Activations in the left fronto-parietal cortex, left putamen and left thalamus in conjunction with right cerebellar activations suggest that the left cortico-thalamo-cerebellar loop may be a dominant loop. Dynamic causal modeling (DCM) indicates the presence of causal influences from the left to the right cerebral cortex. In conclusion, melody-playing with bimanual MSM is a complex in-phase task and may help activate the bilateral cortical areas, and left hemisphere is dominant according to laterality indices and DCM results. On the other hand, bimanual rhythm is a simpler in-phase task and may help activate subcortical areas, which might be independent of the voluntary cortical task.
View details for DOI 10.1097/WNR.0000000000001433
View details for PubMedID 32221114