Dr. Gatidis completed his medical training at the University of Tuebingen / Germany and received his Diploma in Mathematics from from the Universities of Tuebingen and Hagen / Germany. His research is focused on multiparametric oncologic medical imaging including hybrid imaging as well as on methods and applications of machine learning for medical image analysis.
Associate Professor - University Medical Line, Radiology - Pediatric Radiology
Board Certification: District Medical Association of South Wurttemberg, Radiology (2017)
Residency: Eberhard-Karls-Univ Tubingen/Germany Germany
Medical Education: Eberhard-Karls-Univ Tubingen/Germany (2011) Germany
Medical Education, University of Tuebingen / Germany, Medicine (2011)
Board Certification, Southern German Medical Association, Diagnostic Radiology (2017)
Residency, University Hospital Tuebingen / Germany, Diagnostic Radiology (2017)
Clinical Text Summarization: Adapting Large Language Models Can Outperform Human Experts.
Sifting through vast textual data and summarizing key information from electronic health records (EHR) imposes a substantial burden on how clinicians allocate their time. Although large language models (LLMs) have shown immense promise in natural language processing (NLP) tasks, their efficacy on a diverse range of clinical summarization tasks has not yet been rigorously demonstrated. In this work, we apply domain adaptation methods to eight LLMs, spanning six datasets and four distinct clinical summarization tasks: radiology reports, patient questions, progress notes, and doctor-patient dialogue. Our thorough quantitative assessment reveals trade-offs between models and adaptation methods in addition to instances where recent advances in LLMs may not improve results. Further, in a clinical reader study with ten physicians, we show that summaries from our best-adapted LLMs are preferable to human summaries in terms of completeness and correctness. Our ensuing qualitative analysis highlights challenges faced by both LLMs and human experts. Lastly, we correlate traditional quantitative NLP metrics with reader study scores to enhance our understanding of how these metrics align with physician preferences. Our research marks the first evidence of LLMs outperforming human experts in clinical text summarization across multiple tasks. This implies that integrating LLMs into clinical workflows could alleviate documentation burden, empowering clinicians to focus more on personalized patient care and the inherently human aspects of medicine.
View details for DOI 10.21203/rs.3.rs-3483777/v1
View details for PubMedID 37961377
View details for PubMedCentralID PMC10635391
Quantification of intratumoural heterogeneity in mice and patients via machine-learning models trained on PET-MRI data
NATURE BIOMEDICAL ENGINEERING
In oncology, intratumoural heterogeneity is closely linked with the efficacy of therapy, and can be partially characterized via tumour biopsies. Here we show that intratumoural heterogeneity can be characterized spatially via phenotype-specific, multi-view learning classifiers trained with data from dynamic positron emission tomography (PET) and multiparametric magnetic resonance imaging (MRI). Classifiers trained with PET-MRI data from mice with subcutaneous colon cancer quantified phenotypic changes resulting from an apoptosis-inducing targeted therapeutic and provided biologically relevant probability maps of tumour-tissue subtypes. When applied to retrospective PET-MRI data of patients with liver metastases from colorectal cancer, the trained classifiers characterized intratumoural tissue subregions in agreement with tumour histology. The spatial characterization of intratumoural heterogeneity in mice and patients via multimodal, multiparametric imaging aided by machine-learning may facilitate applications in precision oncology.
View details for DOI 10.1038/s41551-023-01047-9
View details for Web of Science ID 001000747500001
View details for PubMedID 37277483
View details for PubMedCentralID 5378302
AI Transformers for Radiation Dose Reduction in Serial Whole-Body PET Scans.
Radiology. Artificial intelligence
2023; 5 (3): e220246
To develop a deep learning approach that enables ultra-low-dose, 1% of the standard clinical dosage (3 MBq/kg), ultrafast whole-body PET reconstruction in cancer imaging.In this Health Insurance Portability and Accountability Act-compliant study, serial fluorine 18-labeled fluorodeoxyglucose PET/MRI scans of pediatric patients with lymphoma were retrospectively collected from two cross-continental medical centers between July 2015 and March 2020. Global similarity between baseline and follow-up scans was used to develop Masked-LMCTrans, a longitudinal multimodality coattentional convolutional neural network (CNN) transformer that provides interaction and joint reasoning between serial PET/MRI scans from the same patient. Image quality of the reconstructed ultra-low-dose PET was evaluated in comparison with a simulated standard 1% PET image. The performance of Masked-LMCTrans was compared with that of CNNs with pure convolution operations (classic U-Net family), and the effect of different CNN encoders on feature representation was assessed. Statistical differences in the structural similarity index measure (SSIM), peak signal-to-noise ratio (PSNR), and visual information fidelity (VIF) were assessed by two-sample testing with the Wilcoxon signed rank t test.The study included 21 patients (mean age, 15 years ± 7 [SD]; 12 female) in the primary cohort and 10 patients (mean age, 13 years ± 4; six female) in the external test cohort. Masked-LMCTrans-reconstructed follow-up PET images demonstrated significantly less noise and more detailed structure compared with simulated 1% extremely ultra-low-dose PET images. SSIM, PSNR, and VIF were significantly higher for Masked-LMCTrans-reconstructed PET (P < .001), with improvements of 15.8%, 23.4%, and 186%, respectively.Masked-LMCTrans achieved high image quality reconstruction of 1% low-dose whole-body PET images.Keywords: Pediatrics, PET, Convolutional Neural Network (CNN), Dose Reduction Supplemental material is available for this article. © RSNA, 2023.
View details for DOI 10.1148/ryai.220246
View details for PubMedID 37293349
View details for PubMedCentralID PMC10245181
Local control and patient reported outcomes after online MR guided stereotactic body radiotherapy of liver metastases
FRONTIERS IN ONCOLOGY
2023; 12: 1095633
Stereotactic body radiotherapy (SBRT) is used to treat liver metastases with the intention of ablation. High local control rates were shown. Magnetic resonance imaging guided radiotherapy (MRgRT) provides the opportunity of a marker-less liver SBRT treatment due to the high soft tissue contrast. We report herein on one of the largest cohorts of patients treated with online MRgRT of liver metastases focusing on oncological outcome, toxicity, patient reported outcome measures (PROMs), quality of life.Patients treated for liver metastases with online MR-guided SBRT at a 1,5 T MR-Linac (Unity, Elekta, Crawley, UK) between March 2019 and December 2021 were included in this prospective study. UK SABR guidelines were used for organs at risk constraints. Oncological endpoints such as survival parameters (overall survival, progression-free survival) and local control as well as patient reported acceptance and quality of life data (EORTC QLQ-C30 questionnaire) were assessed. For toxicity scoring the Common Toxicity Criteria Version 5 were used.A total of 51 patients with 74 metastases were treated with a median of five fractions. The median applied BED GTV D98 was 84,1 Gy. Median follow-up was 15 months. Local control of the irradiated liver metastasis after 12 months was 89,6%, local control of the liver was 40,3%. Overall survival (OS) after 12 months was 85.1%. Progression free survival (PFS) after 12 months was 22,4%. Local control of the irradiated liver lesion was 100% after three years when a BED ≥100 Gy was reached. The number of treated lesions did not impact local control neither of the treated or of the hepatic control. Patient acceptance of online MRgSBRT was high. There were no acute grade ≥ 3 toxicities. Quality of life data showed no significant difference comparing baseline and follow-up data.Online MR guided radiotherapy is a noninvasive, well-tolerated and effective treatment for liver metastases. Further prospective trials with the goal to define patients who actually benefit most from an online adaptive workflow are currently ongoing.
View details for DOI 10.3389/fonc.2022.1095633
View details for Web of Science ID 000921644400001
View details for PubMedID 36727060
View details for PubMedCentralID PMC9885175
Low-count whole-body PET/MRI restoration: an evaluation of dose reduction spectrum and five state-of-the-art artificial intelligence models.
European journal of nuclear medicine and molecular imaging
To provide a holistic and complete comparison of the five most advanced AI models in the augmentation of low-dose 18F-FDG PET data over the entire dose reduction spectrum.In this multicenter study, five AI models were investigated for restoring low-count whole-body PET/MRI, covering convolutional benchmarks - U-Net, enhanced deep super-resolution network (EDSR), generative adversarial network (GAN) - and the most cutting-edge image reconstruction transformer models in computer vision to date - Swin transformer image restoration network (SwinIR) and EDSR-ViT (vision transformer). The models were evaluated against six groups of count levels representing the simulated 75%, 50%, 25%, 12.5%, 6.25%, and 1% (extremely ultra-low-count) of the clinical standard 3 MBq/kg 18F-FDG dose. The comparisons were performed upon two independent cohorts - (1) a primary cohort from Stanford University and (2) a cross-continental external validation cohort from Tübingen University - in order to ensure the findings are generalizable. A total of 476 original count and simulated low-count whole-body PET/MRI scans were incorporated into this analysis.For low-count PET restoration on the primary cohort, the mean structural similarity index (SSIM) scores for dose 6.25% were 0.898 (95% CI, 0.887-0.910) for EDSR, 0.893 (0.881-0.905) for EDSR-ViT, 0.873 (0.859-0.887) for GAN, 0.885 (0.873-0.898) for U-Net, and 0.910 (0.900-0.920) for SwinIR. In continuation, SwinIR and U-Net's performances were also discreetly evaluated at each simulated radiotracer dose levels. Using the primary Stanford cohort, the mean diagnostic image quality (DIQ; 5-point Likert scale) scores of SwinIR restoration were 5 (SD, 0) for dose 75%, 4.50 (0.535) for dose 50%, 3.75 (0.463) for dose 25%, 3.25 (0.463) for dose 12.5%, 4 (0.926) for dose 6.25%, and 2.5 (0.534) for dose 1%.Compared to low-count PET images, with near-to or nondiagnostic images at higher dose reduction levels (up to 6.25%), both SwinIR and U-Net significantly improve the diagnostic quality of PET images. A radiotracer dose reduction to 1% of the current clinical standard radiotracer dose is out of scope for current AI techniques.
View details for DOI 10.1007/s00259-022-06097-w
View details for PubMedID 36633614
Self-supervised contrastive learning with random walks for medical image segmentation with limited annotations
COMPUTERIZED MEDICAL IMAGING AND GRAPHICS
2023; 104: 102174
Medical image segmentation has seen significant progress through the use of supervised deep learning. Hereby, large annotated datasets were employed to reliably segment anatomical structures. To reduce the requirement for annotated training data, self-supervised pre-training strategies on non-annotated data were designed. Especially contrastive learning schemes operating on dense pixel-wise representations have been introduced as an effective tool. In this work, we expand on this strategy and leverage inherent anatomical similarities in medical imaging data. We apply our approach to the task of semantic segmentation in a semi-supervised setting with limited amounts of annotated volumes. Trained alongside a segmentation loss in one single training stage, a contrastive loss aids to differentiate between salient anatomical regions that conform to the available annotations. Our approach builds upon the work of Jabri et al. (2020), who proposed cyclical contrastive random walks (CCRW) for self-supervision on palindromes of video frames. We adapt this scheme to operate on entries of paired embedded image slices. Using paths of cyclical random walks bypasses the need for negative samples, as commonly used in contrastive approaches, enabling the algorithm to discriminate among relevant salient (anatomical) regions implicitly. Further, a multi-level supervision strategy is employed, ensuring adequate representations of local and global characteristics of anatomical structures. The effectiveness of reducing the amount of required annotations is shown on three MRI datasets. A median increase of 8.01 and 5.90 pp in the Dice Similarity Coefficient (DSC) compared to our baseline could be achieved across all three datasets in the case of one and two available annotated examples per dataset.
View details for DOI 10.1016/j.compmedimag.2022.102174
View details for Web of Science ID 000923565500001
View details for PubMedID 36640485
- Avoiding Shortcut-Learning by Mutual Information Minimization in Deep Learning-Based Image Processing IEEE ACCESS 2023; 11: 64070-64086
Impact of endorectal filling on interobserver variability of MRI based rectal primary tumor delineation
CLINICAL AND TRANSLATIONAL RADIATION ONCOLOGY
2023; 38: 1-5
Online adaptive MR-guided radiotherapy allows for the reduction of safety margins in dose escalated treatment of rectal tumors. With the use of smaller margins, precise tumor delineation becomes more critical. In the present study we investigated the impact of rectal ultrasound gel filling on interobserver variability in delineation of primary rectal tumor volumes.Six patients with locally advanced rectal cancer were scanned on a 1.5 T MRI-Linac without (MRI_e) and with application of 100 cc of ultrasound gel transanally (MRI_f). Eight international radiation oncologists expert in the treatment of gastrointestinal cancers delineated the gross tumor volume (GTV) on both MRI scans. MRI_f scans were provided to the participating centers after MRI_e scans had been returned. Interobserver variability was analyzed by either comparing the observers' delineations with a reference delineation (approach 1) and by building all possible pairs between observers (approach 2). Dice Similarity Index (DICE) and 95 % Hausdorff-Distance (95 %HD) were calculated.Rectal ultrasound gel filling was well tolerated by all patients. Overall, interobserver agreement was superior in MRI_f scans based on median DICE (0.81 vs 0.74, p < 0.005 for approach 1 and 0.76 vs 0.64, p < 0.0001 for approach 2) and 95 %HD (6.9 mm vs 4.2 mm for approach 1, p = 0.04 and 8.9 mm vs 6.1 mm, p = 0.04 for approach 2). Delineated median tumor volumes and inter-quartile ranges were 26.99 cc [18.01-50.34 cc] in MRI_e and 44.20 [19.72-61.59 cc] in MRI_f scans respectively, p = 0.012.Although limited by the small number of patients, in this study the application of rectal ultrasound gel resulted in higher interobserver agreement in rectal GTV delineation. The endorectal gel filling might be a useful tool for future dose escalation strategies.
View details for DOI 10.1016/j.ctro.2022.09.002
View details for Web of Science ID 000927150300001
View details for PubMedID 36299279
View details for PubMedCentralID PMC9589000
A whole-body FDG-PET/CT Dataset with manually annotated Tumor Lesions.
2022; 9 (1): 601
We describe a publicly available dataset of annotated Positron Emission Tomography/Computed Tomography (PET/CT) studies. 1014 whole body Fluorodeoxyglucose (FDG)-PET/CT datasets (501 studies of patients with malignant lymphoma, melanoma and non small cell lung cancer (NSCLC) and 513 studies without PET-positive malignant lesions (negative controls)) acquired between 2014 and 2018 were included. All examinations were acquired on a single, state-of-the-art PET/CT scanner. The imaging protocol consisted of a whole-body FDG-PET acquisition and a corresponding diagnostic CT scan. All FDG-avid lesions identified as malignant based on the clinical PET/CT report were manually segmented on PET images in a slice-per-slice (3D) manner. We provide the anonymized original DICOM files of all studies as well as the corresponding DICOM segmentation masks. In addition, we provide scripts for image processing and conversion to different file formats (NIfTI, mha, hdf5). Primary diagnosis, age and sex are provided as non-imaging information. We demonstrate how this dataset can be used for deep learning-based automated analysis of PET/CT data and provide the trained deep learning model.
View details for DOI 10.1038/s41597-022-01718-3
View details for PubMedID 36195599
Standardised lesion segmentation for imaging biomarker quantitation: a consensus recommendation from ESR and EORTC
INSIGHTS INTO IMAGING
2022; 13 (1): 159
Lesion/tissue segmentation on digital medical images enables biomarker extraction, image-guided therapy delivery, treatment response measurement, and training/validation for developing artificial intelligence algorithms and workflows. To ensure data reproducibility, criteria for standardised segmentation are critical but currently unavailable.A modified Delphi process initiated by the European Imaging Biomarker Alliance (EIBALL) of the European Society of Radiology (ESR) and the European Organisation for Research and Treatment of Cancer (EORTC) Imaging Group was undertaken. Three multidisciplinary task forces addressed modality and image acquisition, segmentation methodology itself, and standards and logistics. Devised survey questions were fed via a facilitator to expert participants. The 58 respondents to Round 1 were invited to participate in Rounds 2-4. Subsequent rounds were informed by responses of previous rounds.Items with ≥ 75% consensus are considered a recommendation. These include system performance certification, thresholds for image signal-to-noise, contrast-to-noise and tumour-to-background ratios, spatial resolution, and artefact levels. Direct, iterative, and machine or deep learning reconstruction methods, use of a mixture of CE marked and verified research tools were agreed and use of specified reference standards and validation processes considered essential. Operator training and refreshment were considered mandatory for clinical trials and clinical research. Items with a 60-74% agreement require reporting (site-specific accreditation for clinical research, minimal pixel number within lesion segmented, use of post-reconstruction algorithms, operator training refreshment for clinical practice). Items with ≤ 60% agreement are outside current recommendations for segmentation (frequency of system performance tests, use of only CE-marked tools, board certification of operators, frequency of operator refresher training). Recommendations by anatomical area are also specified.
View details for DOI 10.1186/s13244-022-01287-4
View details for Web of Science ID 000863839900001
View details for PubMedID 36194301
View details for PubMedCentralID PMC9532485
Artificial Intelligence in Oncological Hybrid Imaging
ROFO-FORTSCHRITTE AUF DEM GEBIET DER RONTGENSTRAHLEN UND DER BILDGEBENDEN VERFAHREN
2023; 195 (02): 105-114
Artificial intelligence (AI) applications have become increasingly relevant across a broad spectrum of settings in medical imaging. Due to the large amount of imaging data that is generated in oncological hybrid imaging, AI applications are desirable for lesion detection and characterization in primary staging, therapy monitoring, and recurrence detection. Given the rapid developments in machine learning (ML) and deep learning (DL) methods, the role of AI will have significant impact on the imaging workflow and will eventually improve clinical decision making and outcomes.The first part of this narrative review discusses current research with an introduction to artificial intelligence in oncological hybrid imaging and key concepts in data science. The second part reviews relevant examples with a focus on applications in oncology as well as discussion of challenges and current limitations.AI applications have the potential to leverage the diagnostic data stream with high efficiency and depth to facilitate automated lesion detection, characterization, and therapy monitoring to ultimately improve quality and efficiency throughout the medical imaging workflow. The goal is to generate reproducible, structured, quantitative diagnostic data for evidence-based therapy guidance in oncology. However, significant challenges remain regarding application development, benchmarking, and clinical implementation.· Hybrid imaging generates a large amount of multimodality medical imaging data with high complexity and depth.. · Advanced tools are required to enable fast and cost-efficient processing along the whole radiology value chain.. · AI applications promise to facilitate the assessment of oncological disease in hybrid imaging with high quality and efficiency for lesion detection, characterization, and response assessment. The goal is to generate reproducible, structured, quantitative diagnostic data for evidence-based oncological therapy guidance.. · Selected applications in three oncological entities (lung, prostate, and neuroendocrine tumors) demonstrate how AI algorithms may impact imaging-based tasks in hybrid imaging and potentially guide clinical decision making..· Feuerecker B, Heimer M, Geyer T et al. Artificial Intelligence in Oncological Hybrid Imaging. Fortschr Röntgenstr 2023; 195: 105 - 114.
View details for DOI 10.1055/a-1909-7013
View details for Web of Science ID 000860856200002
View details for PubMedID 36170852
- Antibody-Guided Molecular Imaging of Aspergillus Lung Infections in Leukemia Patients JOURNAL OF NUCLEAR MEDICINE 2022; 63 (9): 1450-1451
Self-supervised learning for automated anatomical tracking in medical image data with minimal human labeling effort
COMPUTER METHODS AND PROGRAMS IN BIOMEDICINE
2022; 225: 107085
Tracking of anatomical structures in time-resolved medical image data plays an important role for various tasks such as volume change estimation or treatment planning. State-of-the-art deep learning techniques for automated tracking, while providing accurate results, require large amounts of human-labeled training data making their wide-spread use time- and resource-intensive. Our contribution in this work is the implementation and adaption of a self-supervised learning (SSL) framework that addresses this bottleneck of training data generation.To this end we adapted and implemented an SSL framework that allows for automated anatomical tracking without the necessity for human-labeled training data. We evaluated this method by comparison to conventional- and deep learning optical flow (OF)-based tracking methods. We applied all methods on three different time-resolved medical image datasets (abdominal MRI, cardiac MRI, and echocardiography) and assessed their accuracy regarding tracking of pre-defined anatomical structures within and across individuals.We found that SSL-based tracking as well as OF-based methods provide accurate results for simple, rigid and smooth motion patterns. However, regarding more complex motion, e.g. non-rigid or discontinuous motion patterns in the cardiac region, and for cross-subject anatomical matching, SSL-based tracking showed markedly superior performance.We conclude that automated tracking of anatomical structures on time-resolved medical image data with minimal human labeling effort is feasible using SSL and can provide superior results compared to conventional and deep learning OF-based methods.
View details for DOI 10.1016/j.cmpb.2022.107085
View details for Web of Science ID 000878991900011
View details for PubMedID 36044801
Combination of Whole-Body Baseline CT Radiomics and Clinical Parameters to Predict Response and Survival in a Stage-IV Melanoma Cohort Undergoing Immunotherapy
2022; 14 (12)
This study investigated whether a machine-learning-based combination of radiomics and clinical parameters was superior to the use of clinical parameters alone in predicting therapy response after three months, and overall survival after six and twelve months, in stage-IV malignant melanoma patients undergoing immunotherapy with PD-1 checkpoint inhibitors and CTLA-4 checkpoint inhibitors.A random forest model using clinical parameters (demographic variables and tumor markers = baseline model) was compared to a random forest model using clinical parameters and radiomics (extended model) via repeated 5-fold cross-validation. For this purpose, the baseline computed tomographies of 262 stage-IV malignant melanoma patients treated at a tertiary referral center were identified in the Central Malignant Melanoma Registry, and all visible metastases were three-dimensionally segmented (n = 6404).The extended model was not significantly superior compared to the baseline model for survival prediction after six and twelve months (AUC (95% CI): 0.664 (0.598, 0.729) vs. 0.620 (0.545, 0.692) and AUC (95% CI): 0.600 (0.526, 0.667) vs. 0.588 (0.481, 0.629), respectively). The extended model was not significantly superior compared to the baseline model for response prediction after three months (AUC (95% CI): 0.641 (0.581, 0.700) vs. 0.656 (0.587, 0.719)).The study indicated a potential, but non-significant, added value of radiomics for six-month and twelve-month survival prediction of stage-IV melanoma patients undergoing immunotherapy.
View details for DOI 10.3390/cancers14122992
View details for Web of Science ID 000819643900001
View details for PubMedID 35740659
View details for PubMedCentralID PMC9221470
1.5T MR Linac for MR-guided stereotactic Radiotherapy of Patients with HCC: first clinical Results
SPRINGER HEIDELBERG. 2022: S134
View details for Web of Science ID 000791788000293
Prospective Toxicity and Outcome Analysis after MR-guided stereotactic Radiotherapy (MRgSBRT) of Liver Metastases at 1.5 T MR Linac
SPRINGER HEIDELBERG. 2022: S37-S38
View details for Web of Science ID 000791788000070
Inter-Observer Variability of MR-based Contouring of Rectal Cancer by MTAs compared to Radiation Oncologists
SPRINGER HEIDELBERG. 2022: S112
View details for Web of Science ID 000791788000243
Prediction of clinical Complete Remission by using functional MR Imaging during Chemoradiation in Rectal Cancer
SPRINGER HEIDELBERG. 2022: S58
View details for Web of Science ID 000791788000113
Interobserver Variability in Primary Tumor Contouring of Rectal Cancer to dose-escalated Radiotherapy
SPRINGER HEIDELBERG. 2022: S32
View details for Web of Science ID 000791788000059
Hybrid Cardiac Magnetic Resonance/Fluorodeoxyglucose Positron Emission Tomography to Differentiate Active From Chronic Cardiac Sarcoidosis
2022; 15 (3): 445-456
The purpose of this study was to investigate the diagnostic value of simultaneous hybrid cardiac magnetic resonance (CMR) and 18F-fluorodeoxyglucose positron emission tomography (FDG-PET) for detection and differentiation of active (aCS) from chronic (cCS) cardiac sarcoidosis.Late gadolinium enhancement (LGE) CMR and FDG-PET are both established imaging techniques for the detection of CS. However, there are limited data regarding the value of a comprehensive simultaneous hybrid CMR/FDG-PET imaging approach that includes CMR mapping techniques.Forty-three patients with biopsy-proven extracardiac sarcoidosis (median age: 48 years, interquartile range: 37-57 years, 65% male) were prospectively enrolled for evaluation of suspected CS. After dietary preparation for suppression of myocardial glucose metabolism, patients were evaluated on a 3-T hybrid PET/MR scanner. The CMR protocol included T1 and T2 mapping, myocardial function, and LGE imaging. We assumed aCS if PET and CMR (ie, LGE or T1/T2 mapping) were both positive (PET+/CMR+), cCS if PET was negative but CMR was positive (PET-/CMR+), and no CS if patients were CMR negative regardless of PET findings.Among the 43 patients, myocardial glucose uptake was suppressed successfully in 36 (84%). Hybrid CMR/FDG-PET revealed aCS in 13 patients (36%), cCS in 5 (14%), and no CS in 18 (50%). LGE was present in 14 patients (39%); T1 mapping was abnormal in 10 (27%) and T2 mapping abnormal in 2 (6%). CS was diagnosed based on abnormal T1 mapping in 4 out of 18 CS patients (22%) who were LGE negative. PET FDG uptake was present in 17 (47%) patients.Comprehensive simultaneous hybrid CMR/FDG-PET imaging is useful for the detection of CS and provides additional value for identifying active disease. Our results may have implications for enhanced diagnosis as well as improved identification of patients with aCS in whom anti-inflammatory therapy may be most beneficial.
View details for DOI 10.1016/j.jcmg.2021.08.018
View details for Web of Science ID 000782633300003
View details for PubMedID 34656480
- Real Time Landmark Detection for Within- and Cross Subject Tracking With Minimal Human Supervision IEEE ACCESS 2022; 10: 81192-81202
LAPNet: Non-Rigid Registration Derived in k-Space for Magnetic Resonance Imaging
IEEE TRANSACTIONS ON MEDICAL IMAGING
2021; 40 (12): 3686-3697
Physiological motion, such as cardiac and respiratory motion, during Magnetic Resonance (MR) image acquisition can cause image artifacts. Motion correction techniques have been proposed to compensate for these types of motion during thoracic scans, relying on accurate motion estimation from undersampled motion-resolved reconstruction. A particular interest and challenge lie in the derivation of reliable non-rigid motion fields from the undersampled motion-resolved data. Motion estimation is usually formulated in image space via diffusion, parametric-spline, or optical flow methods. However, image-based registration can be impaired by remaining aliasing artifacts due to the undersampled motion-resolved reconstruction. In this work, we describe a formalism to perform non-rigid registration directly in the sampled Fourier space, i.e. k-space. We propose a deep-learning based approach to perform fast and accurate non-rigid registration from the undersampled k-space data. The basic working principle originates from the Local All-Pass (LAP) technique, a recently introduced optical flow-based registration. The proposed LAPNet is compared against traditional and deep learning image-based registrations and tested on fully-sampled and highly-accelerated (with two undersampling strategies) 3D respiratory motion-resolved MR images in a cohort of 40 patients with suspected liver or lung metastases and 25 healthy subjects. The proposed LAPNet provided consistent and superior performance to image-based approaches throughout different sampling trajectories and acceleration factors.
View details for DOI 10.1109/TMI.2021.3096131
View details for Web of Science ID 000724511900038
View details for PubMedID 34242163
Detection of lung lesions in breath-hold VIBE and free-breathing Spiral VIBE MRI compared to CT
INSIGHTS INTO IMAGING
2021; 12 (1): 175
Detection of pulmonary nodules in MRI requires fast imaging strategies without respiratory motion impairment, such as single-breath-hold Cartesian VIBE. As patients with pulmonary diseases have limited breath-hold capacities, this study investigates the clinical feasibility of non-Cartesian Spiral VIBE under free-breathing compared to CT as the gold standard.Prospective analysis of 27 oncological patients examined in PET/CT and PET/MR. A novel motion-robust 3D ultrashort-echo-time (UTE) MR sequence was evaluated in comparison with CT and conventional breath-hold MR. CT scans were performed under breath-hold in end-expiratory and end-inspiratory position (CT ex, CT in). MR data was acquired with non-contrast-enhanced breath-hold Cartesian VIBE followed by a free-breathing 3D UTE Spiral VIBE. Impact of respiratory motion on pulmonary evaluation was investigated by two readers in Cartesian VIBE, followed by UTE Spiral VIBE and CT ex and the reference standard of CT in. Diagnostic accuracy was calculated, and visual image quality assessed.Higher detection rate and sensitivity of pulmonary nodules in free-breathing UTE Spiral VIBE in comparison with breath-hold Cartesian VIBE were found for lesions > 10 mm (UTE Spiral VIBE/VIBE/CT ex): 93%/54%/100%; Lesions 5-10 mm: 67%/25%/ 92%; Lesions < 5 mm: 11%/11%/78%. Lobe-based analysis revealed sensitivities and specificities of 64%/96%/41% and 96%/93%/100% for UTE Spiral VIBE/VIBE/CT ex.Free-breathing UTE Spiral VIBE indicates higher sensitivity for detection of pulmonary nodules than breath-hold Cartesian VIBE and is a promising but time-consuming approach. However, sensitivity and specificity of inspiratory CT remain superior in comparison and should be preferred for detection of pulmonary lesions.
View details for DOI 10.1186/s13244-021-01124-0
View details for Web of Science ID 000722208600002
View details for PubMedID 34817715
View details for PubMedCentralID PMC8613318
MRI-Derived Radiomics Features of Hepatic Fat Predict Metabolic States in Individuals without Cardiovascular Disease
2021; 28: S1-S10
To investigate radiomics features of hepatic fat as potential biomarkers of type 2 diabetes mellitus (T2DM) and metabolic syndrome (MetS) in individuals without overt cardiovascular disease, and benchmarking against hepatic proton density fat fraction (PDFF) and the body mass index (BMI).This study collected liver radiomics features of 310 individuals that were part of a case-controlled imaging substudy embedded in a prospective cohort. Individuals had known T2DM (n = 39; 12.6 %) and MetS (n = 107; 34.5 %) status, and were divided into stratified training (n = 232; 75 %) and validation (n = 78; 25 %) sets. Six hundred eighty-four MRI radiomics features were extracted for each liver volume of interest (VOI) on T1-weighted dual-echo Dixon relative fat water content (rfwc) maps. Test-retest and inter-rater variance was simulated by additionally extracting radiomics features using noise augmented rfwc maps and deformed volume of interests. One hundred and seventy-one features with test-retest reliability (ICC(1,1)) and inter-rater agreement (ICC(3,k)) of ≥0.85 on the training set were considered stable. To construct predictive random forest (RF) models, stable features were filtered using univariate RF analysis followed by sequential forward aggregation. The predictive performance was evaluated on the independent validation set with area under the curve of the receiver operating characteristic (AUROC) and balanced accuracy (AccuracyB).On the validation set, the radiomics RF models predicted T2DM with AUROC of 0.835 and AccuracyB of 0.822 and MetS with AUROC of 0.838 and AccuracyB of 0.787, outperforming the RF models trained on the benchmark parameters PDFF and BMI.Hepatic radiomics features may serve as potential imaging biomarkers for T2DM and MetS.
View details for DOI 10.1016/j.acra.2020.06.030
View details for Web of Science ID 000719455800001
View details for PubMedID 32800693
Effectiveness of Chest CT in Children: CT Findings in Relation to the Clinical Question
ROFO-FORTSCHRITTE AUF DEM GEBIET DER RONTGENSTRAHLEN UND DER BILDGEBENDEN VERFAHREN
2022; 194 (03): 281-290
To estimate the effectiveness and efficiency of chest CT in children based on the suspected diagnosis in relation to the number of positive, negative, and inconclusive CT results.In this monocentric retrospective study at a university hospital with a division of pediatric radiology, 2019 chest CT examinations (973 patients; median age: 10.5 years; range: 2 days to 17.9 years) were analyzed with regards to clinical data, including the referring department, primary questions or suspected diagnosis, and CT findings. It was identified if the clinical question was answered, whether the suspected diagnosis was confirmed or ruled out, and if additional findings (clinically significant or minor) were detected.The largest clinical subgroup was the hematooncological subgroup (n = 987), with frequent questions for inflammation/pneumonia (66 % in this subgroup). Overall, CT provided conclusive results in 97.6 % of all scans. In 1380 scans (70 %), the suspected diagnosis was confirmed. In 406/2019 cases (20 %), the CT scan was negative also in terms of an additional finding. In 8 of 9 clinical categories, the proportion of positive results was over 50 %. There were predominantly negative results (110/179; 61 %) in pre-stem cell transplant evaluation. In the subgroup of trauma management, 81/144 exams (57 %) showed positive results, including combined injuries (n = 23). 222/396 (56 %) of all additional findings were estimated to be clinically significant.In a specialized center, the effectiveness of pediatric chest CT was excellent when counting the conclusive results. However, to improve efficiency, the clinical evaluation before imaging appears crucial to prevent unnecessary CT examinations.· Pediatric chest CT in specialized centers has a high diagnostic value.. · CT identifies relevant changes besides the working hypothesis in clinically complex situations.. · Pre-CT clinical evaluation is crucial, especially in the context of suspected pneumonia..· Esser M, Tsiflikas I, Kraus MS et al. Effectiveness of Chest CT in Children: CT Findings in Relation to the Clinical Question. Fortschr Röntgenstr 2022; 194: 281 - 290.
View details for DOI 10.1055/a-1586-3023
View details for Web of Science ID 000707988800004
View details for PubMedID 34649290
PET/MRI Improves Management of Children with Cancer.
Journal of nuclear medicine : official publication, Society of Nuclear Medicine
2021; 62 (10): 1334-1340
Integrated PET/MRI has shown significant clinical value for staging and restaging of children with cancer by providing functional and anatomic tumor evaluation with a 1-stop imaging test and with up to 80% reduced radiation exposure compared with 18F-FDG PET/CT. This article reviews clinical applications of 18F-FDG PET/MRI that are relevant for pediatric oncology, with particular attention to the value of PET/MRI for patient management. Early adopters from 4 different institutions share their insights about specific advantages of PET/MRI technology for the assessment of young children with cancer. We discuss how whole-body PET/MRI can be of value in the evaluation of certain anatomic regions, such as soft tissues and bone marrow, as well as specific PET/MRI interpretation hallmarks in pediatric patients. We highlight how whole-body PET/MRI can improve the clinical management of children with lymphoma, sarcoma, and neurofibromatosis, by reducing the number of radiologic examinations needed (and consequently the radiation exposure), without losing diagnostic accuracy. We examine how PET/MRI can help in differentiating malignant tumors versus infectious or inflammatory diseases. Future research directions toward the use of PET/MRI for treatment evaluation of patients undergoing immunotherapy and assessment of different theranostic agents are also briefly explored. Lessons learned from applications in children might also be extended to evaluations of adult patients.
View details for DOI 10.2967/jnumed.120.259747
View details for PubMedID 34599010
A novel approach for radiotherapy dose escalation in rectal cancer using online MR-guidance and rectal ultrasound gel filling-Rationale and first in human
RADIOTHERAPY AND ONCOLOGY
2021; 164: 37-42
Dose escalated radiotherapy has previously been investigated as a strategy to increase complete response rates in rectal cancer. However large safety margins are required using cone-beam computed tomography guided radiotherapy leading to high doses to organs at risk or insufficient target volume coverage in order to keep dose constraints. We herein present the first clinical application of a new technique for dose escalation in rectal cancer using online magnetic resonance (MR)-guidance and rectal ultrasound gel filling.A 73-year-old patient with distal cT3a cN0 cM0 rectal cancer was referred for definitive radiochemotherapy with the goal of organ preservation after multidisciplinary discussion. A dose of 45 Gy in 25 fractions with a stereotactic integrated boost to the primary tumor of 50 Gy with concomitant 5-fluorouracil was prescribed. Furthermore, a boost to the primary tumor with 3 Gy per fraction using the adapt-to-shape workflow on a 1.5 T MR-Linac was planned once weekly. For the boost fractions 100 cc of ultrasound gel was applied rectally in order to improve tumor visibility and distancing of uninvolved rectal mucosa. In order to determine the required planning target volume margin diagnostic scans of ten rectal cancer patients conducted with rectal ultrasound gel filling were studied.Based on the ten diagnostic scans an average isotropic margin of 4 mm was found to be sufficient to cover 95% of the target volume during an online adaptive workflow. Three boost fractions were applied, mean treatment duration was 22:34 min. Treatment was well tolerated by the patient with no more than PRO-CTCAE grade I° toxicity of any kind. The rectal ultrasound gel filling resulted in superior visibility of the tumor and reduced the dose to the involved mucosa especially in the high dose range compared with a boost plan calculated without any filling. A considerable tumor shrinkage was observed during treatment from 17.43 cc at baseline to 4 cc in week four.This novel method appears to be a simple but effective strategy for dose escalated radiotherapy in rectal cancer. Based on the encouraging observation, a prospective trial is currently under preparation.
View details for DOI 10.1016/j.radonc.2021.09.002
View details for Web of Science ID 000708441100001
View details for PubMedID 34534612
Overcoming barriers to data sharing with medical image generation: a comprehensive evaluation
NPJ DIGITAL MEDICINE
2021; 4 (1): 141
Privacy concerns around sharing personally identifiable information are a major barrier to data sharing in medical research. In many cases, researchers have no interest in a particular individual's information but rather aim to derive insights at the level of cohorts. Here, we utilise generative adversarial networks (GANs) to create medical imaging datasets consisting entirely of synthetic patient data. The synthetic images ideally have, in aggregate, similar statistical properties to those of a source dataset but do not contain sensitive personal information. We assess the quality of synthetic data generated by two GAN models for chest radiographs with 14 radiology findings and brain computed tomography (CT) scans with six types of intracranial haemorrhages. We measure the synthetic image quality by the performance difference of predictive models trained on either the synthetic or the real dataset. We find that synthetic data performance disproportionately benefits from a reduced number of classes. Our benchmark also indicates that at low numbers of samples per class, label overfitting effects start to dominate GAN training. We conducted a reader study in which trained radiologists discriminate between synthetic and real images. In accordance with our benchmark results, the classification accuracy of radiologists improves with an increasing resolution. Our study offers valuable guidelines and outlines practical conditions under which insights derived from synthetic images are similar to those that would have been derived from real data. Our results indicate that synthetic data sharing may be an attractive alternative to sharing real patient-level data in the right setting.
View details for DOI 10.1038/s41746-021-00507-3
View details for Web of Science ID 000698951100001
View details for PubMedID 34561528
View details for PubMedCentralID PMC8463544
Targeting extracellular and juxtamembrane FGFR2 mutations in chemotherapy-refractory cholangiocarcinoma
NPJ PRECISION ONCOLOGY
2021; 5 (1): 80
Intrahepatic cholangiocarcinoma (iCCA) has emerged as a promising candidate for precision medicine, especially in the case of activating FGFR2 gene fusions. In addition to fusions, a considerable fraction of iCCA patients reveals FGFR2 mutations, which might lead to uncontrolled activation of the FGFR2 pathway but are mostly of unknown functional significance. A current challenge for molecular tumor boards (MTB) is to predict the functional consequences of such FGFR2 alterations to guide potential treatment decisions. We report two iCCA patients with extracellular and juxtamembrane FGFR2 mutations. After in silico investigation of the alterations and identification of activated FGFR2 downstream targets in tumor specimens by immunohistochemistry and transcriptome analysis, the MTB recommended treatment with an FGFR-inhibiting tyrosine kinase inhibitor. Both patients developed a rapidly detectable and prolonged partial response to treatment. These two cases suggest an approach to characterize further detected FGFR2 mutations in iCCA to enable patients´ selection for a successful application of the FGFR -inhibiting drugs.
View details for DOI 10.1038/s41698-021-00220-0
View details for Web of Science ID 000694237900001
View details for PubMedID 34480077
View details for PubMedCentralID PMC8417271
Dynamic whole-body F-18-FDG PET/CT in patients with unclear lung tumors - evaluation of multiparametric dynamic imaging in a clinical setting
SPRINGER. 2021: S479
View details for Web of Science ID 000709355001337
Weakly supervised segmentation of tumor lesions in PET-CT hybrid imaging
JOURNAL OF MEDICAL IMAGING
2021; 8 (5): 054003
Purpose: We introduce and evaluate deep learning methods for weakly supervised segmentation of tumor lesions in whole-body fluorodeoxyglucose-positron emission tomography (FDG-PET) based solely on binary global labels ("tumor" versus "no tumor"). Approach: We propose a three-step approach based on (i) a deep learning framework for image classification, (ii) subsequent generation of class activation maps (CAMs) using different CAM methods (CAM, GradCAM, GradCAM++, ScoreCAM), and (iii) final tumor segmentation based on the aforementioned CAMs. A VGG-based classification neural network was trained to distinguish between PET image slices with and without FDG-avid tumor lesions. Subsequently, the CAMs of this network were used to identify the tumor regions within images. This proposed framework was applied to FDG-PET/CT data of 453 oncological patients with available manually generated ground-truth segmentations. Quantitative segmentation performance was assessed for the different CAM approaches and compared with the manual ground truth segmentation and with supervised segmentation methods. In addition, further biomarkers (MTV and TLG) were extracted from the segmentation masks. Results: A weakly supervised segmentation of tumor lesions was feasible with satisfactory performance [best median Dice score 0.47, interquartile range (IQR) 0.35] compared with a fully supervised U-Net model (median Dice score 0.72, IQR 0.36) and a simple threshold based segmentation (Dice score 0.29, IQR 0.28). CAM, GradCAM++, and ScoreCAM yielded similar results. However, GradCAM led to inferior results (median Dice score: 0.12, IQR 0.21) and was likely to ignore multiple instances within a given slice. CAM, GradCAM++, and ScoreCAM yielded accurate estimates of metabolic tumor volume (MTV) and tumor lesion glycolysis. Again, worse results were observed for GradCAM. Conclusions: This work demonstrated the feasibility of weakly supervised segmentation of tumor lesions and accurate estimation of derived metrics such as MTV and tumor lesion glycolysis.
View details for DOI 10.1117/1.JMI.8.5.054003
View details for Web of Science ID 000720829600019
View details for PubMedID 34660843
View details for PubMedCentralID PMC8510879
Uncertainty estimation and explainability in deep learning-based age estimation of the human brain: Results from the German National Cohort MRI study
COMPUTERIZED MEDICAL IMAGING AND GRAPHICS
2021; 92: 101967
Brain ageing is a complex neurobiological process associated with morphological changes that can be assessed on MRI scans. Recently, Deep learning (DL)-based approaches have been proposed for the prediction of chronological brain age from MR images yielding high accuracy. These approaches, however, usually do not address quantification of uncertainty and, therefore, intrinsic physiological variability. Considering uncertainty is essential for the interpretation of the difference between predicted and chronological age. In addition, DL-based models lack in explainability compared to classical approaches like voxel-based morphometry. In this study, we aim to address both, modeling uncertainty and providing visual explanations to explore physiological patterns in brain ageing. T1-weighted brain MRI datasets of 10691 participants of the German National Cohort Study, drawn from the general population, were included in this study (chronological age from 20 to 72 years). A regression model based on a 3D Convolutional Neural Network taking into account aleatoric noise was implemented for global as well as regional brain age estimation. We observed high overall accuracy of global brain age estimation with a mean absolute error of 3.2 ± 2.5 years and mean uncertainty of 2.9 ± 0.6 years. Regional brain age estimation revealed higher estimation accuracy and lower uncertainty in central compared to peripheral brain regions. Visual explanations illustrating the importance of brain sub-regions were generated using Grad-CAM: the derived saliency maps showed a high relevance of the lateral and third ventricles, the insular lobe as well as parts of the basal ganglia and the internal capsule.
View details for DOI 10.1016/j.compmedimag.2021.101967
View details for Web of Science ID 000694676100001
View details for PubMedID 34392229
Lymph Node Staging with a Combined Protocol of F-18-FDG PET/MRI and Sentinel Node SPECT/CT: A Prospective Study in Patients with FIGO I/II Cervical Carcinoma
JOURNAL OF NUCLEAR MEDICINE
2021; 62 (8): 1062-1067
Lymph node metastasis (LNM) is present in a minority of patients with early stages of cervical carcinomas. As conventional imaging including PET/CT has shown limited sensitivity, systematic lymphadenectomies are often conducted for staging purposes. Therefore, the aim of this prospective study was to analyze the impact of 18F-FDG PET/MRI in addition to sentinel lymph node (SLN) biopsy on lymph node (LN) status. Methods: Forty-two women with an initial diagnosis of Fédération Internationale de Gynécologie et d'Obstétrique (FIGO) IA-IIB cervical carcinoma were included between March 2016 and April 2019. Each patient underwent preoperative whole-body 18F-FDG PET/MRI and SLN imaging with SPECT/CT after intracervical injection of 99mTc-labeled nanocolloid. Systematic lymphadenectomy and SLN biopsy served as the reference standard. Staging using PET/MRI was performed by nuclear medicine and radiology experts working in consensus. Results: One patient was excluded from surgical staging because of liver metastases newly diagnosed on PET/MRI. The overall prevalence of LNM in the remaining 41 patients was 29.3% (12/41). Five of 12 patients with LNM had solely small metastases with a maximum diameter of 5 mm. The consensus interpretation showed PET/MRI to have a specificity of 100% (29/29; 95% CI, 88.3%-100%) for LNM staging but a low sensitivity, 33.3% (4/12; 95% CI, 12.8%-60.9%). LN size was the most important factor for the detectability of metastases, since only LNMs larger than 5 mm could be identified by PET/MRI (sensitivity, 57.1% for >5 mm and 0% for ≤5 mm). Paraaortic LNM was evaluated accurately in 3 of the 4 patients with paraaortic LN metastasis. SLNs were detectable by SPECT/CT in 82.9% of the patients or 69.0% of the hemipelves. In cases with an undetectable SLN on SPECT/CT, the malignancy rate was considerably higher (31.2% vs. 19.3%). The combination of PET/MRI and SLN SPECT/CT improved the detection of pelvic LNM from 33.3% to 75%. Conclusion:18F-FDG PET/MRI is a highly specific N-staging method and improves LNM detection. Because of the limited sensitivity in frequently occurring small LNMs, PET/MRI should be combined with SLN mapping. The proposed combined protocol helps to decide whether extensive surgical staging is necessary in patients with FIGO I/II cervical cancer.
View details for DOI 10.2967/jnumed.120.255919
View details for Web of Science ID 000711570000014
View details for PubMedID 33509973
View details for PubMedCentralID PMC8833872
F-18-Fluoride PET/CT Imaging of Medication-Related Osteonecrosis of the Jaw in Conservative Treatment-A Case Report
FRONTIERS IN ONCOLOGY
2021; 11: 700397
Medication-related osteonecrosis of the jaw (MRONJ) is a serious side effect in antiresorptive treatment. Treatment of MRONJ is considered primarily conservative with oral mouth rinses and antibiotics but may demand surgery, depending on the complaints and general condition of the patient, the extent of the necrosis, and the overall prognosis with respect to the underlying disease. A 77 year old female patient with invasive ductal breast cancer and bone metastases was treated with intravenous bisphosphonate (BP) zoledronic acid. During therapy, she developed MRONJ in the mandible with severe pain. Clinical examination revealed confluent exposed bone of the lower left jaw and a fistula at the right molar region. The panoramic radiograph revealed a mandibular osseous involvement with diffuse radiopaque areas between radiolucent areas. For preoperative planning, 18F-fluoride positron emission tomography/computed tomography (PET/CT) of the jaw was performed, showing substantially increased 18F-fluoride uptake in regions 38 to 47 of the mandible with a focal gap in region 36 (area of clinically exposed bone). CT revealed medullary sclerosis and cortical thickening with confluent periosteal reaction and focal cortical erosion in the regions 37 to 42, whereas the regions 43 to 47 were only subtly sclerotic without cortical thickening. After systemic antibiotic therapy with sultamicillin following significant symptom and pain relief, 18F-fluoride PET/CT imaging was performed again after 5 months. No changes in either CT and PET were observed in regions 38 to 42, whereas the bony sclerosis was slightly increased in regions 43 to 47 with a slight reduction of 18F-fluoride uptake. 18F-fluoride PET/CT showed no significant changes assessing the extent of MRONJ prior and after systemic antibiotic therapy, providing no evidence that conservative treatment reduced the extent of the MRONJ-affected jawbone. The additional information of 18F-fluoride PET enables to identify the true extent of MRONJ which may be underestimated by CT imaging alone. Patients with MRONJ undergoing conservative treatment could benefit because additional imaging may be avoided as the pre-therapeutic 18F-fluoride PET/CT delivers all information needed for further treatment. Our findings support the recommendation of a surgical approach as long-term antibiotics cannot downsize the extent of MRONJ.
View details for DOI 10.3389/fonc.2021.700397
View details for Web of Science ID 000673662300001
View details for PubMedID 34277447
View details for PubMedCentralID PMC8281890
Age-Net: An MRI-Based Iterative Framework for Brain Biological Age Estimation
IEEE TRANSACTIONS ON MEDICAL IMAGING
2021; 40 (7): 1778-1791
The concept of biological age (BA) - although important in clinical practice - is hard to grasp mainly due to the lack of a clearly defined reference standard. For specific applications, especially in pediatrics, medical image data are used for BA estimation in a routine clinical context. Beyond this young age group, BA estimation is mostly restricted to whole-body assessment using non-imaging indicators such as blood biomarkers, genetic and cellular data. However, various organ systems may exhibit different aging characteristics due to lifestyle and genetic factors. Thus, a whole-body assessment of the BA does not reflect the deviations of aging behavior between organs. To this end, we propose a new imaging-based framework for organ-specific BA estimation. In this initial study we focus mainly on brain MRI. As a first step, we introduce a chronological age (CA) estimation framework using deep convolutional neural networks (Age-Net). We quantitatively assess the performance of this framework in comparison to existing state-of-the-art CA estimation approaches. Furthermore, we expand upon Age-Net with a novel iterative data-cleaning algorithm to segregate atypical-aging patients (BA [Formula: see text] CA) from the given population. We hypothesize that the remaining population should approximate the true BA behavior. We apply the proposed methodology on a brain magnetic resonance image (MRI) dataset containing healthy individuals as well as Alzheimer's patients with different dementia ratings. We demonstrate the correlation between the predicted BAs and the expected cognitive deterioration in Alzheimer's patients. A statistical and visualization-based analysis has provided evidence regarding the potential and current challenges of the proposed methodology.
View details for DOI 10.1109/TMI.2021.3066857
View details for Web of Science ID 000668842500004
View details for PubMedID 33729932
1.5 T MR-linac Planning Study comparing two different Rectum-Boost Strategies in Patients with locally advanced Rectal Cancer
SPRINGER HEIDELBERG. 2021: S151-S152
View details for Web of Science ID 000664126100271
Deep Learning-Based Automated Abdominal Organ Segmentation in the UK Biobank and German National Cohort Magnetic Resonance Imaging Studies
2021; 56 (6): 401-408
The aims of this study were to train and evaluate deep learning models for automated segmentation of abdominal organs in whole-body magnetic resonance (MR) images from the UK Biobank (UKBB) and German National Cohort (GNC) MR imaging studies and to make these models available to the scientific community for analysis of these data sets.A total of 200 T1-weighted MR image data sets of healthy volunteers each from UKBB and GNC (400 data sets in total) were available in this study. Liver, spleen, left and right kidney, and pancreas were segmented manually on all 400 data sets, providing labeled ground truth data for training of a previously described U-Net-based deep learning framework for automated medical image segmentation (nnU-Net). The trained models were tested on all data sets using a 4-fold cross-validation scheme. Qualitative analysis of automated segmentation results was performed visually; performance metrics between automated and manual segmentation results were computed for quantitative analysis. In addition, interobserver segmentation variability between 2 human readers was assessed on a subset of the data.Automated abdominal organ segmentation was performed with high qualitative and quantitative accuracy on UKBB and GNC data. In more than 90% of data sets, no or only minor visually detectable qualitative segmentation errors occurred. Mean Dice scores of automated segmentations compared with manual reference segmentations were well higher than 0.9 for the liver, spleen, and kidneys on UKBB and GNC data and around 0.82 and 0.89 for the pancreas on UKBB and GNC data, respectively. Mean average symmetric surface distance was between 0.3 and 1.5 mm for the liver, spleen, and kidneys and between 2 and 2.2 mm for pancreas segmentation. The quantitative accuracy of automated segmentation was comparable with the agreement between 2 human readers for all organs on UKBB and GNC data.Automated segmentation of abdominal organs is possible with high qualitative and quantitative accuracy on whole-body MR imaging data acquired as part of UKBB and GNC. The results obtained and deep learning models trained in this study can be used as a foundation for automated analysis of thousands of MR data sets of UKBB and GNC and thus contribute to tackling topical and original scientific questions.
View details for DOI 10.1097/RLI.0000000000000755
View details for Web of Science ID 000650001200008
View details for PubMedID 33930003
- The International Radiomics Platform - An Initiative of the German and Austrian Radiological Societies (vol 193, pg 276, 2021) ROFO-FORTSCHRITTE AUF DEM GEBIET DER RONTGENSTRAHLEN UND DER BILDGEBENDEN VERFAHREN 2021; 193 (03): E3
Case Report: Combined CDK4/6 and MEK Inhibition in Refractory CDKN2A and NRAS Mutant Melanoma
FRONTIERS IN ONCOLOGY
2021; 11: 643156
There are only limited treatment options for metastatic NRAS mutant melanoma patients with resistance to immune checkpoint inhibitors. Besides activation of the mitogen-activated protein (MAP) kinase pathway, they often have additional disturbances in cell cycle regulation. However, unlike BRAF mutant melanoma, no targeted therapy has yet been approved for NRAS mutant melanoma so far. Here we present a NRAS mutant melanoma patient with response to combined binimetinib and ribociclib therapy following characterization of the molecular defects of the tumor by panel sequencing. Next generation sequencing (708 cancer genes) of a soft tissue metastasis revealed a homozygous deletion of CDKN2A in addition to the previously known NRAS mutation, as well as amplification of CCNE1 and CDK6. Immunohistochemical staining of the altered cell cycle genes confirmed loss of p16, reduced expression of p21 and high expression of CDK6 and cyclin D1. As the patient had been progressive on combined immunotherapy, targeted therapy with combined MEK and CDK4/6 inhibition was initiated as recommended by the molecular tumor board. Response to treatment was monitored with PET/CT and liquid biopsy, serum LDH, and S100. In addition, a patient-derived xenograft (PDX) was used to prove the efficacy of the two drugs in combination. Furthermore, senescence-associated beta-galactosidase staining showed that more cells were senescent under the combination treatment of binimetinib and ribociclib. Our case demonstrates how an individualized, molecular-based therapeutic approach could be found based on next-generation sequencing results. Furthermore our report highlights the fruitful and efficient collaboration of dermatooncologists, human geneticists, molecular pathologists, biochemists, radiologists, and nuclear physicians. Further studies are urgently needed to expand the very limited therapeutic landscape of NRAS mutated melanoma.
View details for DOI 10.3389/fonc.2021.643156
View details for Web of Science ID 000628892500001
View details for PubMedID 33732653
View details for PubMedCentralID PMC7959243
Artificial intelligence enables whole-body positron emission tomography scans with minimal radiation exposure.
European journal of nuclear medicine and molecular imaging
PURPOSE: To generate diagnostic 18F-FDG PET images of pediatric cancer patients from ultra-low-dose 18F-FDG PET input images, using a novel artificial intelligence (AI) algorithm.METHODS: We used whole-body 18F-FDG-PET/MRI scans of 33 children and young adults with lymphoma (3-30years) to develop a convolutional neural network (CNN), which combines inputs from simulated 6.25% ultra-low-dose 18F-FDG PET scans and simultaneously acquired MRI scans to produce a standard-dose 18F-FDG PET scan. The image quality of ultra-low-dose PET scans, AI-augmented PET scans, and clinical standard PET scans was evaluated by traditional metrics in computer vision and by expert radiologists and nuclear medicine physicians, using Wilcoxon signed-rank tests and weighted kappa statistics.RESULTS: The peak signal-to-noise ratio and structural similarity index were significantly higher, and the normalized root-mean-square error was significantly lower on the AI-reconstructed PET images compared to simulated 6.25% dose images (p<0.001). Compared to the ground-truth standard-dose PET, SUVmax values of tumors and reference tissues were significantly higher on the simulated 6.25% ultra-low-dose PET scans as a result of image noise. After the CNN augmentation, the SUVmax values were recovered to values similar to the standard-dose PET. Quantitative measures of the readers' diagnostic confidence demonstrated significantly higher agreement between standard clinical scans and AI-reconstructed PET scans (kappa=0.942) than 6.25% dose scans (kappa=0.650).CONCLUSIONS: Our CNN model could generate simulated clinical standard 18F-FDG PET images from ultra-low-dose inputs, while maintaining clinically relevant information in terms of diagnostic accuracy and quantitative SUV measurements.
View details for DOI 10.1007/s00259-021-05197-3
View details for PubMedID 33527176
Impact of PET/CT on management of patients with esophageal cancer-results from a PET/CT registry study
EUROPEAN JOURNAL OF RADIOLOGY
2021; 136: 109524
To investigate the impact of positron emission tomography/computed tomography (PET/CT) on clinical management in patients with esophageal cancer and its link to overall survival (OS) in a real-world setting.A patient cohort with advanced esophageal cancer undergoing PET/CT was prospectively enrolled in a registry study between 04/2013 and 06/2019. Intended patient management prior and after PET/CT was documented based on standardized questionnaire data. Management changes after PET/CT were recorded including major changes concerning the treatment goal (curative vs. palliative) and minor changes (therapy adjustments). OS was analyzed for subgroups with squamous cell carcinomas (SCC) or adenocarcinomas (AC) and stratified for extent of metastatic disease and treatment goals.257 patients (53 female;65.5 ± 10.0 yr.) were included. After PET/CT, major changes of intended therapy were observed in 34/257 patients (13.2%), from curative to palliative (8.2%), palliative to curative (1.9%) and from "not finally determined" to a curative (1.9%) or palliative (1.2%) concept. Minor changes were found in 62/257 patients (24.1%). Invasive procedures and additional imaging were intended in 70/257 (27.2%) and 94/257 (36.6%) patients before PET/CT and 20/257 (7.8%) and 8/257 (3.1%) patients after PET/CT. Curative therapy concepts based on PET/CT were associated with a longer OS (3.5 yr.[95%CI 3.1-3.8 yr.]) as compared to palliative concepts (0.9 yr.[95%CI 0.6-1.2 yr.];p < 0.0001). Patients with SCC had a worse prognosis (2.4 yr.[95%CI 2.0-2.9 yr.]) as compared to patients with AC (3.2 yr.[95%CI 2.7-3.7 yr.];p = 0.01).In patients with advanced esophageal cancer, PET/CT has a significant impact on clinical management by improving the selection of individualized treatment strategies and avoiding additional diagnostic procedures.
View details for DOI 10.1016/j.ejrad.2021.109524
View details for Web of Science ID 000620374400013
View details for PubMedID 33434862
- Uncertainty-Guided Progressive GANs for Medical Image Translation SPRINGER INTERNATIONAL PUBLISHING AG. 2021: 614-624
Marker-less online MR-guided stereotactic body radiotherapy of liver metastases at a 1.5 T MR-Linac - Feasibility, workflow data and patient acceptance
CLINICAL AND TRANSLATIONAL RADIATION ONCOLOGY
2021; 26: 55-61
Stereotactic body radiotherapy (SBRT) is an established ablative treatment for liver tumors with excellent local control rates. Magnetic resonance imaging guided radiotherapy (MRgRT) provides superior soft tissue contrast and may therefore facilitate a marker-less liver SBRT workflow. The goal of the present study was to investigate feasibility, workflow parameters, toxicity and patient acceptance of MRgSBRT on a 1.5 T MR-Linac.Ten consecutive patients with liver metastases treated on a 1.5 T MR-Linac were included in this prospective trial. Tumor delineation was performed on four-dimensional computed tomography scans and both exhale triggered and free-breathing T2 MRI scans from the MR-Linac. An internal target volume based approach was applied. Organ at risk constraints were based on the UKSABR guidelines (Version 6.1). Patient acceptance regarding device specific aspects was assessed and toxicity was scored according to the common toxicity criteria of adverse events, version 5.Nine of ten tumors were clearly visible on the 1.5 T MR-Linac. No patient had fiducial markers placed for treatment. All patients were treated with three or five fractions. Median dose to 98% of the gross tumor volume was 38.5 Gy. The median time from "patient identity check" until "beam-off" was 31 min. Median beam on time was 9.6 min. Online MRgRT was well accepted in general and no treatment had to be interrupted on patient request. No event of symptomatic radiation induced liver disease was observed after a median follow-up of ten month (range 3-17 months).Our early experience suggests that online 1.5 T MRgSBRT of liver metastases represents a promising new non-invasive marker-free treatment modality based on high image quality, clinically reasonable in-room times and high patient acceptance. Further studies are necessary to assess clinical outcome, to validate advanced motion management and to explore the benefit of online response adaptive liver SBRT.
View details for DOI 10.1016/j.ctro.2020.11.014
View details for Web of Science ID 000600587800009
View details for PubMedID 33319073
View details for PubMedCentralID PMC7723999
1.5 T MR-linac planning study to compare two different strategies of rectal boost irradiation
CLINICAL AND TRANSLATIONAL RADIATION ONCOLOGY
2021; 26: 86-91
To compare treatment plans of two different rectal boost strategies: up-front versus adaptive boost at the 1.5 T MR-Linac.Patients with locally advanced rectal cancer (LARC) underwent standard neoadjuvant radiochemotherapy with 50.4 Gy in 28 fractions. T2-weighted MRI prior and after the treatment session were acquired to contour gross tumor volumes (GTVs) and organs at risk (OARs). The datasets were used to simulate four different boost strategies (all with 15 Gy/5 fractions in addition to 50.4 Gy): up-front boost (5 daily fractions in the first week of treatment) and an adaptive boost (one boost fraction per week). Both strategies were planned using standard and reduced PTV margins. Intra-fraction motion was assessed by pre- and post-treatment MRI-based contours.Five patients were included and a total of 44 MRI sets were evaluated. The median PTV volumes of the adaptive boost were significantly smaller than for the up-front boost (81.4 cm3 vs 44.4 cm3 for PTV with standard margins; 31.2 cm3 vs 15 cm3 for PTV with reduced margins; p = 0.031). With reduced margins the rectum was significantly better spared with an adaptive boost rather than with an up-front boost: V60Gy and V65Gy were 41.2% and 24.8% compared with 59% and 29.9%, respectively (p = 0.031). Median GTV intra-fractional motion was 2 mm (range 0-8 mm).The data suggest that the adaptive boost strategy exploiting tumor-shrinkage and reduced margin might result in better sparing of rectum and anal canal. Individual margin assessment, motion management and real-time adaptive radiotherapy appear attractive applications of the 1.5 T MR-Linac for further testing of individualized and safe dose escalation in patients with rectal cancer.
View details for DOI 10.1016/j.ctro.2020.11.016
View details for Web of Science ID 000600587800013
View details for PubMedID 33336086
View details for PubMedCentralID PMC7732969
Feature-based Response Prediction to Immunotherapy of late-stage Melanoma Patients Using PET/MR Imaging
IEEE. 2021: 1229-1233
View details for Web of Science ID 000632622300247
- UNCERTAINTY-BASED BIOLOGICAL AGE ESTIMATION OF BRAIN MRI SCANS IEEE. 2021: 1100-1104
- AUTOMATED MULTI-ORGAN SEGMENTATION IN PET IMAGES USING CASCADED TRAINING OF A 3D U-NET AND CONVOLUTIONAL AUTOENCODER IEEE. 2021: 1145-1149
The International Radiomics Platform - An Initiative of the German and Austrian Radiological Societies First Application Examples
ROFO-FORTSCHRITTE AUF DEM GEBIET DER RONTGENSTRAHLEN UND DER BILDGEBENDEN VERFAHREN
2021; 193 (03): 276-287
The DRG-ÖRG IRP (Deutsche Röntgengesellschaft-Österreichische Röntgengesellschaft international radiomics platform) represents a web-/cloud-based radiomics platform based on a public-private partnership. It offers the possibility of data sharing, annotation, validation and certification in the field of artificial intelligence, radiomics analysis, and integrated diagnostics. In a first proof-of-concept study, automated myocardial segmentation and automated myocardial late gadolinum enhancement (LGE) detection using radiomic image features will be evaluated for myocarditis data sets. The DRG-ÖRP IRP can be used to create quality-assured, structured image data in combination with clinical data and subsequent integrated data analysis and is characterized by the following performance criteria: Possibility of using multicentric networked data, automatically calculated quality parameters, processing of annotation tasks, contour recognition using conventional and artificial intelligence methods and the possibility of targeted integration of algorithms. In a first study, a neural network pre-trained using cardiac CINE data sets was evaluated for segmentation of PSIR data sets. In a second step, radiomic features were applied for segmental detection of LGE of the same data sets, which were provided multicenter via the IRP. First results show the advantages (data transparency, reliability, broad involvement of all members, continuous evolution as well as validation and certification) of this platform-based approach. In the proof-of-concept study, the neural network demonstrated a Dice coefficient of 0.813 compared to the expert's segmentation of the myocardium. In the segment-based myocardial LGE detection, the AUC was 0.73 and 0.79 after exclusion of segments with uncertain annotation.The evaluation and provision of the data takes place at the IRP, taking into account the FAT (fairness, accountability, transparency) and FAIR (findable, accessible, interoperable, reusable) criteria. It could be shown that the DRG-ÖRP IRP can be used as a crystallization point for the generation of further individual and joint projects. The execution of quantitative analyses with artificial intelligence methods is greatly facilitated by the platform approach of the DRG-ÖRP IRP, since pre-trained neural networks can be integrated and scientific groups can be networked.In a first proof-of-concept study on automated segmentation of the myocardium and automated myocardial LGE detection, these advantages were successfully applied.Our study shows that with the DRG-ÖRP IRP, strategic goals can be implemented in an interdisciplinary way, that concrete proof-of-concept examples can be demonstrated, and that a large number of individual and joint projects can be realized in a participatory way involving all groups. · The DRG-ÖRG IRP is a web/cloud-based radiomics platform based on a public-private partnership.. · The DRG-ÖRG IRP can be used for the creation of quality-assured, structured image data in combination with clinical data and subsequent integrated data analysis.. · First results show the applicability of left ventricular myocardial segmentation using a neural network and segment-based LGE detection using radiomic image features.. · The DRG-ÖRG IRP offers the possibility of integrating pre-trained neural networks and networking of scientific groups..· Overhoff D, Kohlmann P, Frydrychowicz A et al. The International Radiomics Platform - An Initiative of the German and Austrian Radiological Societies. Fortschr Röntgenstr 2021; 193: 276 - 287.
View details for DOI 10.1055/a-1244-2775
View details for Web of Science ID 000592809200002
View details for PubMedID 33242898
Serial DWI in HNC treated on a 1.5 T MR-Linac and benchmark to a reference 3 T diagnostic MR-scanner
ELSEVIER IRELAND LTD. 2020: S927-S928
View details for Web of Science ID 000648572703205
Fully Automated Segmentation and Shape Analysis of the Thoracic Aorta in Non-contrast-enhanced Magnetic Resonance Images of the German National Cohort Study
JOURNAL OF THORACIC IMAGING
2020; 35 (6): 389-398
The purpose of this study was to develop and validate a deep learning-based framework for automated segmentation and vessel shape analysis on non-contrast-enhanced magnetic resonance (MR) data of the thoracic aorta within the German National Cohort (GNC) MR study.One hundred data sets acquired in the GNC MR study were included (56 men, average age 53 y [22 to 72 y]). All participants had undergone non-contrast-enhanced MR imaging of the thoracic vessels. Automated vessel segmentation of the thoracic aorta was performed using a Convolutional Neural Network in a supervised setting with manually annotated data sets as the ground truth. Seventy data sets were used for training; 30 data sets were used for quantitative and qualitative evaluation. Automated shape analysis based on centerline extraction from segmentation masks was performed to derive a diameter profile of the vessel. For comparison, 2 radiologists measured vessel diameters manually.Overall, automated aortic segmentation was successful, providing good qualitative analyses with only minor irregularities in 29 of 30 data sets. One data set with severe MR artifacts led to inadequate automated segmentation results. The mean Dice score of automated vessel segmentation was 0.85. Automated aortic diameter measurements were similar to manual measurements (average difference -0.9 mm, limits of agreement: -5.4 to 3.9 mm), with minor deviations in the order of the interreader agreement between the 2 radiologists (average difference -0.5 mm, limits of agreement: -5.8 to 4.8 mm).Automated segmentation and shape analysis of the thoracic aorta is feasible with high accuracy on non-contrast-enhanced MR imaging using the proposed deep learning approach.
View details for DOI 10.1097/RTI.0000000000000522
View details for Web of Science ID 000583403600013
View details for PubMedID 32349056
Noninvasive, longitudinal imaging-based analysis of body adipose tissue and water composition in a melanoma mouse model and in immune checkpoint inhibitor-treated metastatic melanoma patients
CANCER IMMUNOLOGY IMMUNOTHERAPY
2021; 70 (5): 1263-1275
As cancer cachexia (CC) is associated with cancer progression, early identification would be beneficial. The aim of this study was to establish a workflow for automated MRI-based segmentation of visceral (VAT) and subcutaneous adipose tissue (SCAT) and lean tissue water (LTW) in a B16 melanoma animal model, monitor diseases progression and transfer the protocol to human melanoma patients for therapy assessment.For in vivo monitoring of CC B16 melanoma-bearing and healthy mice underwent longitudinal three-point DIXON MRI (days 3, 12, 17 after subcutaneous tumor inoculation). In a prospective clinical study, 18 metastatic melanoma patients underwent MRI before, 2 and 12 weeks after onset of checkpoint inhibitor therapy (CIT; n = 16). We employed an in-house MATLAB script for automated whole-body segmentation for detection of VAT, SCAT and LTW.B16 mice exhibited a CC phenotype and developed a reduced VAT volume compared to baseline (B16 - 249.8 µl, - 25%; controls + 85.3 µl, + 10%, p = 0.003) and to healthy controls. LTW was increased in controls compared to melanoma mice. Five melanoma patients responded to CIT, 7 progressed, and 6 displayed a mixed response. Responding patients exhibited a very limited variability in VAT and SCAT in contrast to others. Interestingly, the LTW was decreased in CIT responding patients (- 3.02% ± 2.67%; p = 0.0034) but increased in patients with progressive disease (+ 1.97% ± 2.19%) and mixed response (+ 4.59% ± 3.71%).MRI-based segmentation of fat and water contents adds essential additional information for monitoring the development of CC in mice and metastatic melanoma patients during CIT or other treatment approaches.
View details for DOI 10.1007/s00262-020-02765-8
View details for Web of Science ID 000583119500001
View details for PubMedID 33130917
View details for PubMedCentralID PMC8053172
Fully Automated and Standardized Segmentation of Adipose Tissue Compartments via Deep Learning in 3D Whole-Body MRI of Epidemiologic Cohort Studies
2020; 2 (6): e200010
To enable fast and reliable assessment of subcutaneous and visceral adipose tissue compartments derived from whole-body MRI.Quantification and localization of different adipose tissue compartments derived from whole-body MR images is of high interest in research concerning metabolic conditions. For correct identification and phenotyping of individuals at increased risk for metabolic diseases, a reliable automated segmentation of adipose tissue into subcutaneous and visceral adipose tissue is required. In this work, a three-dimensional (3D) densely connected convolutional neural network (DCNet) is proposed to provide robust and objective segmentation. In this retrospective study, 1000 cases (average age, 66 years ± 13 [standard deviation]; 523 women) from the Tuebingen Family Study database and the German Center for Diabetes research database and 300 cases (average age, 53 years ± 11; 152 women) from the German National Cohort (NAKO) database were collected for model training, validation, and testing, with transfer learning between the cohorts. These datasets included variable imaging sequences, imaging contrasts, receiver coil arrangements, scanners, and imaging field strengths. The proposed DCNet was compared to a similar 3D U-Net segmentation in terms of sensitivity, specificity, precision, accuracy, and Dice overlap.Fast (range, 5-7 seconds) and reliable adipose tissue segmentation can be performed with high Dice overlap (0.94), sensitivity (96.6%), specificity (95.1%), precision (92.1%), and accuracy (98.4%) from 3D whole-body MRI datasets (field of view coverage, 450 × 450 × 2000 mm). Segmentation masks and adipose tissue profiles are automatically reported back to the referring physician.Automated adipose tissue segmentation is feasible in 3D whole-body MRI datasets and is generalizable to different epidemiologic cohort studies with the proposed DCNet.Supplemental material is available for this article.© RSNA, 2020.
View details for DOI 10.1148/ryai.2020200010
View details for Web of Science ID 000826480100008
View details for PubMedID 33937847
View details for PubMedCentralID PMC8082356
Simultaneous whole-body PET/MRI with integrated multiparametric MRI for primary staging of high-risk prostate cancer
WORLD JOURNAL OF UROLOGY
2020; 38 (10): 2513-2521
Whole-body positron emission tomography/magnetic resonance imaging (wbPET/MRI) is a promising diagnostic tool of recurrent prostate cancer (PC), but its role in primary staging of high-risk PC (hrPC) is not well defined. Thus, the aim was to compare the diagnostic accuracy for T-staging of PET-blinded reading (PBR) and PET/MRI.In this prospective study, hrPC patients scheduled to radical prostatectomy (RPx) with extended lymphadenectomy (eLND) were staged with wbPET/MRI and either 68Ga-PSMA-11 or 11C-choline including simultaneous multiparametric MRI (mpMRI). Images were assessed in two sessions, first as PBR (mpMRI and wbMRI) and second as wbPET/MRI. Prostate Imaging Reporting and Data System criteria (PIRADS v2) were used for T-staging. Results were correlated with the exact anatomical localization and extension as defined by histopathology. Diagnostic accuracy of cTNM stage according to PBR was compared to pathological pTNM stage as reference standard.Thirty-four patients underwent wbPET/MRI of 68Ga-PSMA-11 (n = 17) or 11C-choline (n = 17). Twenty-four patients meeting the inclusion criteria of localized disease ± nodal disease based on imaging results underwent RPx and eLND, whereas ten patients were excluded from analysis due to metastatic disease. T-stage was best defined by mpMRI with underestimation of tumor lesion size by PET for both tracers. N-stage yielded a per patient sensitivity/specificity comparable to PBR.MpMRI is the primary modality for T-staging in hrPC as PET underestimated T-stage in direct comparison to final pathology. In this selected study, cohort MRI shows no inferiority compared to wbPET/MRI considering N-staging.
View details for DOI 10.1007/s00345-019-03066-1
View details for Web of Science ID 000571851500019
View details for PubMedID 31907632
View details for PubMedCentralID 26742998
Structured reporting in oncologic hybrid imaging: a consensus recommendation
2020; 59 (04): 288-293
Since the clinical introduction of PET/CT in the year of 2001 and PET/MRI in the year of 2010, hybrid imaging-guided precision medicine has become an important component of diagnostic algorithms in oncology. The written report represents the primary mode of communication between the referring physician and both the nuclear medicine physician and the radiologist. Reports have considerable impact on patient management and patient outcome, and serve as a legal documentation of the services provided and the expert impression of the interpreting physician. A high-quality hybrid imaging study should result in a likewise high-quality, structured written report which satisfactorily answers the clinical question of the referring physician. In this manuscript, consensus recommendations for structure and content of oncologic hybrid imaging reports and conclusive impressions are provided. Moreover, exemplary structured reports are provided. The recommendations for structured reporting provided in this document should foster further standardization and harmonization of oncologic reports in the context of hybrid imaging. They should also simplify communication with referring physicians and support both acceptance and appreciation of the clinical value of oncologic hybrid imaging. CITATION FORMAT: · Derlin T, Gatidis S, Krause BJ et al. Konsensusempfehlung zur strukturierten Befunderstellung onkologischer PET-Hybridbildgebung. Nuklearmedizin 2020; 59: DOI:10.1055/a-1176-0275.
View details for DOI 10.1055/a-1176-0275
View details for Web of Science ID 000555834600001
View details for PubMedID 32544954
Structured reporting in oncologic hybrid imaging: a consensus recommendation
ROFO-FORTSCHRITTE AUF DEM GEBIET DER RONTGENSTRAHLEN UND DER BILDGEBENDEN VERFAHREN
2020; 192 (8): 739-744
Since the clinical introduction of PET/CT in the year of 2001 and PET/MRI in the year of 2010, hybrid imaging-guided precision medicine has become an important component of diagnostic algorithms in oncology. The written report represents the primary mode of communication between the referring physician and both the nuclear medicine physician and the radiologist. Reports have considerable impact on patient management and patient outcome, and serve as a legal documentation of the services provided and the expert impression of the interpreting physician. A high-quality hybrid imaging study should result in a likewise high-quality, structured written report which satisfactorily answers the clinical question of the referring physician. In this manuscript, consensus recommendations for structure and content of oncologic hybrid imaging reports and conclusive impressions are provided. Moreover, exemplary structured reports are provided. The recommendations for structured reporting provided in this document should foster further standardization and harmonization of oncologic reports in the context of hybrid imaging. They should also simplify communication with referring physicians and support both acceptance and appreciation of the clinical value of oncologic hybrid imaging. CITATION FORMAT: · Derlin T, Gatidis S, Krause BJ et al. Konsensusempfehlung zur strukturierten Befunderstellung onkologischer PET-Hybridbildgebung. Fortschr Röntgenstr 2020; 192: DOI:10.1055/a-1179-6603.
View details for DOI 10.1055/a-1179-6603
View details for Web of Science ID 000590148400002
View details for PubMedID 32544953
Clinical and prognostic value of tumor volumetric parameters in melanoma patients undergoing(18)F-FDG-PET/CT: a comparison with serologic markers of tumor burden and inflammation
2020; 20 (1): 44
To investigate the association of tumor volumetric parameters in melanoma patients undergoing 18F-FDG-PET/CT with serologic tumor markers and inflammatory markers and the role as imaging predictors for overall survival.A patient cohort with advanced melanoma undergoing 18F-FDG-PET/CT for planning metastasectomy between 04/2013 and 01/2015 was retrospectively included. The volumetric PET parameters whole-body MTV and whole-body TLG as well as the standard uptake value (SUV) peak were quantified using 50%-isocontour volumes of interests (VOIs) and then correlated with the serologic parameters lactate dehydrogenase (LDH), S-100 protein, c-reactive protein (CRP) and alkaline phosphatase (AP). PET parameters were dichotomized by their respective medians and correlated with overall survival (OS) after PET/CT. OS was compared between patients with or without metastases and increased or not-increased serologic parameters.One hundred seven patients (52 female; 65 ± 13.1yr.) were included. LDH was strongly associated with MTV (rP = 0.73, p < 0.001) and TLG (rP = 0.62, p < 0.001), and moderately associated with SUVpeak (rP = 0.55, p < 0.001). S-100 protein showed a moderate association with MTV (rP = 0.54, p < 0.001) and TLG (rP = 0.48, p < 0.001) and a weak association with SUVpeak (rP = 0.42, p < 0.001). A strong association was observed between CRP and MTV (rP = 0.66, p < 0.001) and a moderate to weak association between CRP and TLG (rP = 0.53, p < 0.001) and CRP and SUVpeak (rP = 0.45, p < 0.001). For differentiation between patients with or without metastases, receiver operating characteristic (ROC) analysis revealed a cut-off value of 198 U/l for serum LDH (AUC 0.81, sensitivity 0.80, specificity 0.72). Multivariate analysis for OS revealed that both MTV and TLG were strong independent prognostic factors. TLG, MTV and SUVpeak above patient median were accompanied with significantly reduced estimated OS compared to the PET parameters below patient median (e.g. TLG: 37.1 ± 3.2 months vs. 55.9 ± 2.5 months, p < 0.001). Correspondingly, both elevated serum LDH and S-100 protein were accompanied with significantly reduced OS (36.5 ± 4.9 months and 37.9 ± 4.4 months) compared to normal serum LDH (49.2 ± 2.4 months, p = 0.01) and normal S-100 protein (49.0 ± 2.5 months, p = 0.01).Tumor volumetric parameters in 18F-FDG-PET/CT serve as prognostic imaging biomarkers in patients with advanced melanoma which are associated with established serologic tumor markers and inflammatory markers.
View details for DOI 10.1186/s40644-020-00322-1
View details for Web of Science ID 000552723100001
View details for PubMedID 32631431
View details for PubMedCentralID PMC7339397
Independent attenuation correction of whole body [F-18]FDG-PET using a deep learning approach with Generative Adversarial Networks
2020; 10 (1): 53
Attenuation correction (AC) of PET data is usually performed using a second imaging for the generation of attenuation maps. In certain situations however-when CT- or MR-derived attenuation maps are corrupted or CT acquisition solely for the purpose of AC shall be avoided-it would be of value to have the possibility of obtaining attenuation maps only based on PET information. The purpose of this study was to thus develop, implement, and evaluate a deep learning-based method for whole body [18F]FDG-PET AC which is independent of other imaging modalities for acquiring the attenuation map.The proposed method is investigated on whole body [18F]FDG-PET data using a Generative Adversarial Networks (GAN) deep learning framework. It is trained to generate pseudo CT images (CTGAN) based on paired training data of non-attenuation corrected PET data (PETNAC) and corresponding CT data. Generated pseudo CTs are then used for subsequent PET AC. One hundred data sets of whole body PETNAC and corresponding CT were used for training. Twenty-five PET/CT examinations were used as test data sets (not included in training). On these test data sets, AC of PET was performed using the acquired CT as well as CTGAN resulting in the corresponding PET data sets PETAC and PETGAN. CTGAN and PETGAN were evaluated qualitatively by visual inspection and by visual analysis of color-coded difference maps. Quantitative analysis was performed by comparison of organ and lesion SUVs between PETAC and PETGAN.Qualitative analysis revealed no major SUV deviations on PETGAN for most anatomic regions; visually detectable deviations were mainly observed along the diaphragm and the lung border. Quantitative analysis revealed mean percent deviations of SUVs on PETGAN of - 0.8 ± 8.6% over all organs (range [- 30.7%, + 27.1%]). Mean lesion SUVs showed a mean deviation of 0.9 ± 9.2% (range [- 19.6%, + 29.2%]).Independent AC of whole body [18F]FDG-PET is feasible using the proposed deep learning approach yielding satisfactory PET quantification accuracy. Further clinical validation is necessary prior to implementation in clinical routine applications.
View details for DOI 10.1186/s13550-020-00644-y
View details for Web of Science ID 000536942500002
View details for PubMedID 32449036
View details for PubMedCentralID PMC7246235
- Combined positron emission tomography and magnetic resonance imaging (PET/MRI) in children and adolescents MONATSSCHRIFT KINDERHEILKUNDE 2020; 168 (5): 416-426
PET/MRI and genetic intrapatient heterogeneity in head and neck cancers
STRAHLENTHERAPIE UND ONKOLOGIE
2020; 196 (6): 542-551
The relation between functional imaging and intrapatient genetic heterogeneity remains poorly understood. The aim of our study was to investigate spatial sampling and functional imaging by FDG-PET/MRI to describe intrapatient tumour heterogeneity.Six patients with oropharyngeal cancer were included in this pilot study. Two tumour samples per patient were taken and sequenced by next-generation sequencing covering 327 genes relevant in head and neck cancer. Corresponding regions were delineated on pretherapeutic FDG-PET/MRI images to extract apparent diffusion coefficients and standardized uptake values.Samples were collected within the primary tumour (n = 3), within the primary tumour and the involved lymph node (n = 2) as well as within two independent primary tumours (n = 1). Genetic heterogeneity of the primary tumours was limited and most driver gene mutations were found ubiquitously. Slightly increasing heterogeneity was found between primary tumours and lymph node metastases. One private predicted driver mutation within a primary tumour and one in a lymph node were found. However, the two independent primary tumours did not show any shared mutations in spite of a clinically suspected field cancerosis. No conclusive correlation between genetic heterogeneity and heterogeneity of PET/MRI-derived parameters was observed.Our limited data suggest that single sampling might be sufficient in some patients with oropharyngeal cancer. However, few driver mutations might be missed and, if feasible, spatial sampling should be considered. In two independent primary tumours, both lesions should be sequenced. Our data with a limited number of patients do not support the concept that multiparametric PET/MRI features are useful to guide biopsies for genetic tumour characterization.
View details for DOI 10.1007/s00066-020-01606-y
View details for Web of Science ID 000521726000004
View details for PubMedID 32211941
Association between metabolic syndrome and hip osteoarthritis in middle-aged men and women from the general population
2020; 15 (3): e0230185
To investigate the impact of metabolic syndrome and its components on osteoarthritis of the hip joints compared to a healthy cohort in the KORA MRI-study.Randomly selected men and women from the general population were classified as having metabolic syndrome, defined as presence of central obesity plus two of the following four components: elevated blood pressure (BP), elevated fasting glucose, elevated triglycerides (TG) and low HDL-cholesterol (HDL-c), or as controls without metabolic syndrome. Therefore, each subject underwent detailed assessment of waist circumference as well as fasting glucose, systolic and diastolic BP, TG, and HDL-c concentrations as well as a full-body MR scan. MR measurements were performed on a 3 Tesla scanner (Magnetom Skyra, Siemens) including a dual-echo Dixon and a T2 SS-FSE sequence for anatomical structures. In order to quantify osteoarthritis of the hip, assessment was performed by two independent, experienced radiologists for joint gap narrowing, osteophytic lipping and subchondral changes (e.g. sclerosis, pseudocysts). Associations between metabolic syndrome components and hip degeneration were estimated by logistic regression models providing odds ratios.Among 354 included participants (mean age: 56.1 ± 9.2 years; 55.4% male), 119 (34%) had metabolic syndrome, while 235 (66%) were part of the control group. Except for elevated blood glucose (p = 0.02), none of the metabolic syndromes' component was independently associated with osteoarthritis. Multivariable adjusted ORs for osteoarthritis of the right hip were 1.00 (95% CI 0.98;1.03), 1.00 (95% CI 0.99;1.00), 1.01 (95% CI 0.99;1.03), 1.00 (95% CI 0.97;1.04) and 1.01 (95% CI 0.96;1.06), and for the left hip 1.00 (95% CI 0.98;1.03), 1.00 (95% CI 1.00;1.01), 1.01 (95% CI 0.99;1.03), 0.99 (95% CI 0.96;1.02) and 1.04 (95% CI 0.99;1.09) for waist circumference, triglyceride, HDL-c and systolic and diastolic BP, respectively. Blood glucose was a borderline non-dependent factor for osteoarthritis of the right hip (OR: 1.02 (95% CI 1.0;1.04); p = 0.05). Furthermore, the compound metabolic syndrome was not significantly associated (OR left hip: 1.53 (95% CI 0.8;2.92), p = 0.20; OR right hip: 1.33 (95% CI 0.72;2.45), p = 0.37) with osteoarthritis of the hip joint. Age as well as gender (left hip) were the only parameters in univariate and multivariate analysis to be significantly associated with osteoarthritis of the hip joint.The compound metabolic syndrome showed no association with osteoarthritis of the hip joint. Age was the only parameter to be dependently and independently associated to osteoarthritis of both hip joints, while elevated blood glucose was independently associated with degeneration of the right hip joint.
View details for DOI 10.1371/journal.pone.0230185
View details for Web of Science ID 000535278500071
View details for PubMedID 32155212
View details for PubMedCentralID PMC7064195
Effects of simulated dose variation on contrast-enhanced CT-based radiomic analysis for Non-Small Cell Lung Cancer
EUROPEAN JOURNAL OF RADIOLOGY
2020; 124: 108804
To examine the potential effect of CT dose variation on radiomic features in vivo using simulated contrast-enhanced CT dose reduction in patients with non-small lung cell cancer (NSCLC).In this retrospective study, we included 69 patients (25 females, 44 males, median age 66 years) with histologically proven NSCLC who underwent a whole contrast-enhanced body FDG-PET/CT for primary staging. To simulate different CT dose levels, we used an algorithm to simulate low-dose CT images based on a noise model derived from phantom experiments. The tumor lesions and reference regions in the paraspinal muscle were manually segmented to obtain three-dimensional regions of interest. Radiomic feature extraction was performed using the PyRadiomics toolbox. The median relative feature value deviation was assessed for each feature and each dose level.The mean segmented tumor volume was 340 ml. T-stages of the primary tumors were primarily T3/4. For NSCLCs, the median relative feature value deviation in the lowest dose images varied for the calculated features from 52.2% to -49.5%. In general, dose-dependent deviations of feature values showed a monotonous increase or decrease with decreasing dose levels. Statistical analyses revealed significant differences between the dose levels in 91% of features.We examined the effects of simulated CT dose reduction on the values of radiomic features in primary NSCLC and showed significant deviations of varying degrees in numerous feature classes. This is a theoretical indicator of potential influence under real conditions, which should be taken into account in clinical use.
View details for DOI 10.1016/j.ejrad.2019.108804
View details for Web of Science ID 000512890500006
View details for PubMedID 31926387
Impact of PET/CT on clinical management in patients with cancer of unknown primary-a PET/CT registry study
2020; 30 (3): 1325-1333
To evaluate the impact of PET/CT on clinical management in patients with cancer of unknown primary (CUP).A cohort of patients with CUP undergoing PET/CT was prospectively enrolled in a local PET/CT registry study between April 2013 and June 2018. Questionnaire data from referring physicians on intended patient management prior and after PET/CT were recorded including items on the intended treatment concept and intended additional diagnostics. Changes in management after PET/CT were recorded. Patient outcome of different cohorts was analyzed for overall survival drawn from patient records.One hundred fifty-five patients (53 female; 63.4 ± 12.1 years) were included. Intended therapeutic management was revised in 45.8% of patients after PET/CT, including major changes affecting the intended treatment goal in 26.5% of patients and minor changes (therapy adjustments) in 19.3% of patients. Invasive and additional diagnostic procedures were intended in 25.8% and 63.2% prior PET/CT and 13.5% and 6.5% after PET/CT. PET/CT-based curative therapy concepts were associated with significantly longer patient survival (4.7 ± 0.3 years) than palliative therapy concepts (1.8 ± 0.5 years, p = .0001). Patients with cervical CUP showed a significantly longer survival (4.3 ± 0.3 years) than patients with extracervical CUP (3.5 ± 0.5 years, p = .01). The identification of the primary did not significantly affect survival.This registry study confirms previous studies reporting that PET/CT significantly influences clinical management in patients with CUP, helping physicians to select a more individualized treatment and to avoid additional diagnostics. Furthermore, we could confirm that tumor localization and extent as shown by PET/CT have a significant impact on patient prognosis.• PET/CT significantly influences intended clinical management in patients with CUP, helping physicians to select a more individualized treatment and to avoid additional diagnostics. • Tumor localization and extent as shown by PET/CT have a significant impact on patient prognosis. • The identification of the primary tumor has no significant impact on overall patient survival.
View details for DOI 10.1007/s00330-019-06518-9
View details for Web of Science ID 000517458800004
View details for PubMedID 31728688
New severity assessment in cystic fibrosis: signal intensity and lung volume compared to LCI and FEV1: preliminary results
2020; 30 (3): 1350-1358
Magnetic resonance imaging (MRI) aids diagnosis in cystic fibrosis (CF) but its use in quantitative severity assessment is under research. This study aims to assess changes in signal intensity (SI) and lung volumes (Vol) during functional MRI and their use as a severity assessment tool in CF patients.The CF intra-hospital standard chest 1.5 T MRI protocol comprises of very short echo-time sequences in submaximal in- and expiration for functional information. Quantitative measurements (Vol/SI at in- and expiration, relative differences (Vol_delta/SI_delta), and cumulative histograms for normalized SI values across the expiratory lung volume) were assessed for correlation to pulmonary function: lung clearance index (LCI) and forced expiratory volume in 1 s (FEV1).In 49 patients (26 male, mean age 17 ± 7 years) significant correlation of Vol_delta and SI_delta (R = 0.86; p < 0.0001) during respiration was observed. Individual cumulated histograms enabled severity disease differentiation (mild, severe) to be visualized (defined by functional parameter: LCI > 10). The expiratory volume at a relative SI of 100% correlated significantly to LCI (R = 0.676 and 0.627; p < 0.0001) and FEV1 (R = - 0.847 and - 0.807; p < 0.0001). Clustering patients according to Vol_SI_100 showed that an amount of ≤ 4% was related to normal, while an amount of > 4% was associated with pathological pulmonary function values.Functional pulmonary MRI provides a radiation-free severity assessment tool and can contribute to early detection of lung impairment in CF. Lung volume with SI below 100% of the inspiratory volume represents overinflated tissue; an amount of 4% of the expiratory lung volume was a relevant turning point.• Signal intensity and lung volumes are used as potential metric parameters for lung impairment. • Quantification of trapped air impacts on therapy management. • Functional pulmonary MRI can contribute to early detection of lung impairment.
View details for DOI 10.1007/s00330-019-06462-8
View details for Web of Science ID 000517458800007
View details for PubMedID 31728685
Paraneoplastic syndrome in undifferentiated embryonic sarcoma of the liver
2020; 10 (1): 11
The undifferentiated embryonic sarcoma of the liver (UESL) is a rare, aggressive tumor mainly affecting children. Since UESL has no specific clinical symptoms or imaging characteristics, many cases of UESL are diagnosed late. The paraneoplastic leukemoid reaction (PLR) is a very rare concomitant of oncological patients associated with poor prognosis. This report describes the clinical course of a patient combining these two rare entities and describes the diagnostic challenges and dynamics of paraneoplastic syndrome.We report a case of UESL in a 46-year-old male who became clinically conspicuous due to pronounced B symptoms. CT and MRI showed a suspicious unifocal liver lesion. As the histological analysis of a tissue sample did not reveal a clear result, an 18F-FDG-PET-CT examination was performed. In addition to a high glucose metabolism of the liver lesion, an increased glucose metabolism in the entire bone marrow was observed. This finding was considered as pronounced paraneoplasia and laparotomy with liver segment resection followed. Immediately after resection of the tumor the paraneoplastic symptoms completely declined and the patient showed no signs of recurrence in the 1-year follow-up.Although UESL is rare and predominantly affects children, this diagnosis should always be considered for unclear unifocal cystic liver lesions, regardless of the patient's age, as appropriate treatment has a good prognosis.
View details for DOI 10.1186/s13550-020-0602-x
View details for Web of Science ID 000514803100001
View details for PubMedID 32072333
View details for PubMedCentralID PMC7028890
Therapy Response Assessment of Pediatric Tumors with Whole-Body Diffusion-weighted MRI and FDG PET/MRI.
Background Whole-body diffusion-weighted (DW) MRI can help detect cancer with high sensitivity. However, the assessment of therapy response often requires information about tumor metabolism, which is measured with fluorine 18 fluorodeoxyglucose (FDG) PET. Purpose To compare tumor therapy response with whole-body DW MRI and FDG PET/MRI in children and young adults. Materials and Methods In this prospective, nonrandomized multicenter study, 56 children and young adults (31 male and 25 female participants; mean age, 15 years ± 4 [standard deviation]; age range, 6-22 years) with lymphoma or sarcoma underwent 112 simultaneous whole-body DW MRI and FDG PET/MRI between June 2015 and December 2018 before and after induction chemotherapy (ClinicalTrials.gov identifier: NCT01542879). The authors measured minimum tumor apparent diffusion coefficients (ADCs) and maximum standardized uptake value (SUV) of up to six target lesions and assessed therapy response after induction chemotherapy according to the Lugano classification or PET Response Criteria in Solid Tumors. The authors evaluated agreements between whole-body DW MRI- and FDG PET/MRI-based response classifications with Krippendorff α statistics. Differences in minimum ADC and maximum SUV between responders and nonresponders and comparison of timing for discordant and concordant response assessments after induction chemotherapy were evaluated with the Wilcoxon test. Results Good agreement existed between treatment response assessments after induction chemotherapy with whole-body DW MRI and FDG PET/MRI (α = 0.88). Clinical response prediction according to maximum SUV (area under the receiver operating characteristic curve = 100%; 95% confidence interval [CI]: 99%, 100%) and minimum ADC (area under the receiver operating characteristic curve = 98%; 95% CI: 94%, 100%) were similar (P = .37). Sensitivity and specificity were 96% (54 of 56 participants; 95% CI: 86%, 99%) and 100% (56 of 56 participants; 95% CI: 54%, 100%), respectively, for DW MRI and 100% (56 of 56 participants; 95% CI: 93%, 100%) and 100% (56 of 56 participants; 95% CI: 54%, 100%) for FDG PET/MRI. In eight of 56 patients who underwent imaging after induction chemotherapy in the early posttreatment phase, chemotherapy-induced changes in tumor metabolism preceded changes in proton diffusion (P = .002). Conclusion Whole-body diffusion-weighted MRI showed significant agreement with fluorine 18 fluorodeoxyglucose PET/MRI for treatment response assessment in children and young adults. © RSNA, 2020 Online supplemental material is available for this article.
View details for DOI 10.1148/radiol.2020192508
View details for PubMedID 32368961
MedGAN: Medical image translation using GANs
COMPUTERIZED MEDICAL IMAGING AND GRAPHICS
2020; 79: 101684
Image-to-image translation is considered a new frontier in the field of medical image analysis, with numerous potential applications. However, a large portion of recent approaches offers individualized solutions based on specialized task-specific architectures or require refinement through non-end-to-end training. In this paper, we propose a new framework, named MedGAN, for medical image-to-image translation which operates on the image level in an end-to-end manner. MedGAN builds upon recent advances in the field of generative adversarial networks (GANs) by merging the adversarial framework with a new combination of non-adversarial losses. We utilize a discriminator network as a trainable feature extractor which penalizes the discrepancy between the translated medical images and the desired modalities. Moreover, style-transfer losses are utilized to match the textures and fine-structures of the desired target images to the translated images. Additionally, we present a new generator architecture, titled CasNet, which enhances the sharpness of the translated medical outputs through progressive refinement via encoder-decoder pairs. Without any application-specific modifications, we apply MedGAN on three different tasks: PET-CT translation, correction of MR motion artefacts and PET image denoising. Perceptual analysis by radiologists and quantitative evaluations illustrate that the MedGAN outperforms other existing translation approaches.
View details for DOI 10.1016/j.compmedimag.2019.101684
View details for Web of Science ID 000514014200007
View details for PubMedID 31812132
UNSUPERVISED ADVERSARIAL CORRECTION OF RIGID MR MOTION ARTIFACTS
IEEE. 2020: 1494-1498
View details for Web of Science ID 000578080300309
- Bayesian Neural Networks for Uncertainty Estimation of Imaging Biomarkers SPRINGER INTERNATIONAL PUBLISHING AG. 2020: 270-280
ipA-MedGAN: INPAINTING OF ARBITRARY REGIONS IN MEDICAL IMAGING
IEEE. 2020: 3005-3009
View details for Web of Science ID 000646178503023
Deep-learning based motion-corrected image reconstruction in 4D magnetic resonance imaging of the body trunk
IEEE. 2020: 976-985
View details for Web of Science ID 000678729400168
Baseline clinical and imaging predictors of treatment response and overall survival of patients with metastatic melanoma undergoing immunotherapy
EUROPEAN JOURNAL OF RADIOLOGY
2019; 121: 108688
We aimed to identify predictive clinical and CT imaging biomarkers and assess their predictive capacity regarding overall survival (OS) and treatment response in patients with metastatic melanoma undergoing immunotherapy.The local institutional ethics committee approved this retrospective study and waived informed patient consent. 103 patients with immunotherapy for metastatic melanoma were randomly divided into training (n = 69) and validation cohort (n = 34). Baseline tumor markers (LDH, S100B), baseline CT imaging biomarkers (tumor burden, Choi density) and CT texture parameters (Entropy, Kurtosis, Skewness, uniformity, MPP, UPP) of the largest target lesion were extracted. To identify treatment response predictors, binary logistic regression analysis was performed in the training cohort and tested in the validation cohort. For OS, Cox regression and Kaplan Maier analyses were performed in the training cohort. Bivariate and multivariate models were established. Goodness of fit was assessed with Harrell's C-index. Potential predictors were tested in the validation cohort also using Cox-regression and Kaplan-Meier analyses.Baseline S100B (Hazard ratio(HR) = 2.543, p0.018), tumor burden (HR = 1.657, p = 0.002) and Kurtosis (HR = 2.484, p < 0.001) were independent predictors of OS and were confirmed in the validation cohort (p < 0.048). Tumor burden and Kurtosis showed incremental predictive capacity allowing a good predictive model when combined with baseline S100B levels (C-index = 0.720). Only S100B was predictive of treatment response (OR ≤ 0.630, p ≤ 0.022). Imaging biomarkers did not predict treatment response.We identified easily obtainable baseline clinical (S100B) and CT predictors (tumor burden and Kurtosis) of OS in patients with metastatic melanoma undergoing immunotherapy. However, imaging predictors did not predict treatment response.
View details for DOI 10.1016/j.ejrad.2019.108688
View details for Web of Science ID 000500465900006
View details for PubMedID 31704599
Assessment of Hepatic Perfusion Using GRASP MRI Bringing Liver MRI on a New Level
2019; 54 (12): 737-743
The aim of this study was to demonstrate the feasibility of hepatic perfusion imaging using dynamic contrast-enhanced (DCE) golden-angle radial sparse parallel (GRASP) magnetic resonance imaging (MRI) for characterizing liver parenchyma and hepatocellular carcinoma (HCC) before and after transarterial chemoembolization (TACE) as a potential alternative to volume perfusion computed tomography (VPCT).Between November 2017 and September 2018, 10 patients (male = 8; mean age, 66.5 ± 8.6 years) with HCC were included in this prospective, institutional review board-approved study. All patients underwent DCE GRASP MRI with high spatiotemporal resolution after injection of liver-specific MR contrast agent before and after TACE. In addition, VPCT was acquired before TACE serving as standard of reference. From the dynamic imaging data of DCE MRI and VPCT, perfusion maps (arterial liver perfusion [mL/100 mL/min], portal liver perfusion [mL/100 mL/min], hepatic perfusion index [%]) were calculated using a dual-input maximum slope model and compared with assess perfusion measures, lesion characteristics, and treatment response using Wilcoxon signed-rank test. To evaluate interreader agreement for measurement repeatability, the interclass correlation coefficient (ICC) was calculated.Perfusion maps could be successfully generated from all DCE MRI and VPCT data. The ICC was excellent for all perfusion maps (ICC ≥ 0.88; P ≤ 0.001). Image analyses revealed perfusion parameters for DCE MRI and VPCT within the same absolute range for tumor and liver tissue. Dynamic contrast-enhanced MRI further enabled quantitative assessment of treatment response showing a significant decrease (P ≤ 0.01) of arterial liver perfusion and hepatic perfusion index in the target lesion after TACE.Dynamic contrast-enhanced GRASP MRI allows for a reliable and robust assessment of hepatic perfusion parameters providing quantitative results comparable to VPCT and enables characterization of HCC before and after TACE, thus posing the potential to serve as an alternative to VPCT.
View details for DOI 10.1097/RLI.0000000000000586
View details for Web of Science ID 000497657600001
View details for PubMedID 31206392
Retrospective correction of motion-affected MR images using deep learning frameworks
MAGNETIC RESONANCE IN MEDICINE
2019; 82 (4): 1527-1540
Motion is 1 extrinsic source for imaging artifacts in MRI that can strongly deteriorate image quality and, thus, impair diagnostic accuracy. In addition to involuntary physiological motion such as respiration and cardiac motion, intended and accidental patient movements can occur. Any impairment by motion artifacts can reduce the reliability and precision of the diagnosis and a motion-free reacquisition can become time- and cost-intensive. Numerous motion correction strategies have been proposed to reduce or prevent motion artifacts. These methods have in common that they need to be applied during the actual measurement procedure with a-priori knowledge about the expected motion type and appearance. For retrospective motion correction and without the existence of any a-priori knowledge, this problem is still challenging.We propose the use of deep learning frameworks to perform retrospective motion correction in a reference-free setting by learning from pairs of motion-free and motion-affected images. For this image-to-image translation problem, we propose and compare a variational auto encoder and generative adversarial network. Feasibility and influences of motion type and optimal architecture are investigated by blinded subjective image quality assessment and by quantitative image similarity metrics.We observed that generative adversarial network-based motion correction is feasible producing near-realistic motion-free images as confirmed by blinded subjective image quality assessment. Generative adversarial network-based motion correction accordingly resulted in images with high evaluation metrics (normalized root mean squared error <0.08, structural similarity index >0.8, normalized mutual information >0.9).Deep learning-based retrospective restoration of motion artifacts is feasible resulting in near-realistic motion-free images. However, the image translation task can alter or hide anatomical features and, therefore, the clinical applicability of this technique has to be evaluated in future studies.
View details for DOI 10.1002/mrm.27783
View details for Web of Science ID 000483917000025
View details for PubMedID 31081955
Independent brain F-18-FDG PET attenuation correction using a deep learning approach with Generative Adversarial Networks
HELLENIC JOURNAL OF NUCLEAR MEDICINE
2019; 22 (3): 179-186
Attenuation correction (AC) of positron emission tomography (PET) data poses a challenge when no transmission data or computed tomography (CT) data are available, e.g. in stand alone PET scanners or PET/magnetic resonance imaging (MRI). In these cases, external imaging data or morphological imaging data are normally used for the generation of attenuation maps. Newly introduced machine learning methods however may allow for direct estimation of attenuation maps from non attenuation-corrected PET data (PETNAC). Our purpose was thus to establish and evaluate a method for independent AC of brain fluorine-18-fluorodeoxyglucose (18F-FDG) PET images only based on PETNAC using Generative Adversarial Networks (GAN).After training of the deep learning GAN framework on a paired training dataset of PETNAC and the corresponding CT images of the head from 50 patients, pseudo-CT images were generated from PETNAC of 40 validation patients, of which 20 were used for technical validation and 20 stemming from patients with CNS disorders were used for clinical validation. Pseudo-CT was used for subsequent AC of these validation data sets resulting in independently attenuation-corrected PET data.Visual inspection revealed a high degree of resemblance of generated pseudo-CT images compared to the acquired CT images in all validation data sets, with minor differences in individual anatomical details. Quantitative analyses revealed minimal underestimation below 5% of standardized uptake value (SUV) in all brain regions in independently attenuation-corrected PET data compared to the reference PET images. Color-coded error maps showed no regional bias and only minimal average errors around ±0%. Using independently attenuation-corrected PET data, no differences in image-based diagnoses were observed in 20 patients with neurological disorders compared to the reference PET images.Independent AC of brain 18F-FDG PET is feasible with high accuracy using the proposed, easy to implement deep learning framework. Further evaluation in clinical cohorts will be necessary to assess the clinical performance of this method.
View details for Web of Science ID 000506147400004
View details for PubMedID 31587027
Comprehensive metabolic and morphologic disease characterization in systemic sclerosis: initial results using combined positron emission tomography and magnetic resonance imaging
QUARTERLY JOURNAL OF NUCLEAR MEDICINE AND MOLECULAR IMAGING
2019; 63 (2): 207-215
The aim of this study was to evaluate the role of metabolic and morphologic parameters derived from simultaneous hybrid PET/MRI in correlation to clinical criteria for an image-based characterization of musculoskeletal, esophagus and lymph node involvement in systemic sclerosis (SSc).Between November 2013 and May 2015, simultaneous whole-body hybrid PET/MRI was performed in 13 prospectively recruited patients with SSc. A mean dose of 241.3 MBq 2-deoxy-2-[18F]fluoro-D-glucose (FDG) was injected. SUVmean and SUVmax values were measured in the spinal bone marrow, spleen, joints, muscles, fasciae, mediastinal lymph nodes and esophagus. MRI abnormalities were scored as 0 (absent), 1 (moderate) and 2 (marked). In addition, organ and skin involvement were graded with clinical sum score (CSS) and modified Rodnan skin score (mRSS), respectively.Results indicate positive correlations between mRSS and fascial FDG-uptake values (fascia summed SUVmax ρ=0.67; fascia summed SUVmean ρ=0.66) that performed better than the MRI sum score (ρ=0.50). Fascial FDG-uptake is also useful in the differentiation between diffuse and limited SSc. Additionally, FDG-PET detected patients with active mediastinal lymphadenopathy and MRI proved to be useful for the delineation of esophagus involvement.Fascial FDG-uptake has a strong correlation with mRSS and can discriminate between limited and diffuse SSc. These results and the detection of active lymphadenopathy and esophagus involvement can identify patients with advanced scleroderma. Combined PET/MRI therefore provides complementary information on the complex pathophysiology and may integrate several imaging procedures in one.
View details for DOI 10.23736/S1824-4785.17.02966-1
View details for Web of Science ID 000486419900012
View details for PubMedID 28478666
Genetic and imaging intratumoral Heterogeneity in Patients with Head and Neck Tumors
SPRINGER HEIDELBERG. 2019: S130
View details for Web of Science ID 000468853300276
State of affairs of hybrid imaging in Europe: two multi-national surveys from 2017
INSIGHTS INTO IMAGING
2019; 10 (1): 57
To assess the current state of hybrid imaging in Europe with respect to operations, reading and reporting, as well as qualification and training.The first survey (LOCAL) was sent to the heads of the departments of radiology and nuclear medicine in Europe in 2017, including 15 questions regarding the organisation of hybrid imaging operations, reporting strategies for PET/CT and the existence of relevant training programmes. The second survey (NATIONAL) consisted of 10 questions and was directed to the national ministries of health of 37 European countries addressing combined training options in radiology and nuclear medicine.In the LOCAL survey, 61 valid responses from 26 European countries were received. In almost half of the institutions, hybrid imaging was performed within a single department, mainly in nuclear medicine departments (31%). In half of the centres (51%), PET/CT reports were performed jointly, while in 20% of the centres, reporting was performed by nuclear medicine physicians. Radiologists were responsible for presenting hybrid imaging results in clinical boards in 34% of responding sites. Integrated hybrid imaging training was available in 41% sites. In the NATIONAL survey, responses from 34 countries were received and demonstrated a heterogeneous landscape of official training possibilities in radiology and nuclear medicine with limited opportunities for additional qualifications in hybrid imaging.The results of these surveys demonstrate a notable heterogeneity in the current practice of hybrid imaging throughout Europe. This heterogeneity exists despite the general consensus that strong professional cooperation is required in order to ensure high clinical quality and to strengthen the clinical role of hybrid imaging.
View details for DOI 10.1186/s13244-019-0741-7
View details for Web of Science ID 000468757300003
View details for PubMedID 31115706
View details for PubMedCentralID PMC6529476
FDG-PET/MRI combined with SPECT/CT guided sentinel lymph node mapping for lymph node staging in patients with early stage cervical and endometrial carcinoma: a prospective study.
SOC NUCLEAR MEDICINE INC. 2019
View details for Web of Science ID 000473116800550
Whole-Body [18F]-FDG-PET/MRI for Oncology: A Consensus Recommendation
ROFO-FORTSCHRITTE AUF DEM GEBIET DER RONTGENSTRAHLEN UND DER BILDGEBENDEN VERFAHREN
2019; 191 (4): 289-297
Combined PET/MR imaging (PET/MRI) was proposed for patient management in 2006 with first commercial versions of integrated whole-body systems becoming available as of 2010. PET/MRI followed the prior evolution of hybrid imaging as attested by the successful adoption of combined PET/CT and SPECT/CT since the early 2000 s. Today, around 150 whole-body PET/MRI systems have become operational worldwide. One of the main application fields of PET/MRI is oncologic imaging. Despite the increasing use of PET/MRI, little governance regarding standardized PET/MRI protocols has been provided to date. Standardization and harmonization of imaging protocols is, however, mandatory for efficient on-site patient management and multi-center studies. This document summarizes consensus recommendations on key aspects of patient referral and preparation, PET/MRI workflow and imaging protocols, as well as reporting strategies for whole-body [18F]-FDG-PET/MRI. These recommendations were created by early adopters and key experts in the field of PET, MRI and PET/MRI. This document is intended to provide guidance for the harmonization and standardization of PET/MRI today and to support wider clinical adoption of this imaging modality for the benefit of patients. CITATION FORMAT: · Umutlu L, Beyer T, Grueneisen JS et al. Whole-Body [18F]-FDG-PET/MRI for Oncology: A Consensus Recommendation. Fortschr Röntgenstr 2019; 191: 289 - 297.
View details for DOI 10.1055/a-0828-8654
View details for Web of Science ID 000463102800003
View details for PubMedID 30818411
Whole-Body [18F]-FDG-PET/MRI for Oncology: A Consensus Recommendation
2019; 58 (2): 68-76
Combined PET/MR imaging (PET/MRI) was proposed for patient management in 2006 with first commercial versions of integrated whole-body systems becoming available as of 2010. PET/MRI followed the prior evolution of hybrid imaging as attested by the successful adoption of combined PET/CT and SPECT/CT since the early 2000 s. Today, around 150 whole-body PET/MRI systems have become operational worldwide. One of the main application fields of PET/MRI is oncologic imaging. Despite the increasing use of PET/MRI, little governance regarding standardized PET/MRI protocols has been provided to date. Standardization and harmonization of imaging protocols is, however, mandatory for efficient on-site patient management and multi-center studies. This document summarizes consensus recommendations on key aspects of patient referral and preparation, PET/MRI workflow and imaging protocols, as well as reporting strategies for whole-body [18F]-FDG-PET/MRI. These recommendations were created by early adopters and key experts in the field of PET, MRI and PET/MRI. This document is intended to provide guidance for the harmonization and standardization of PET/ MRI today and to support wider clinical adoption of this imaging modality for the benefit of patients. CITATION FORMAT:: Umutlu L, Beyer T, Grueneisen JS et al. Whole-Body [18F]-FDG-PET/MRI for Oncology: A Consensus Recommendation. Nuklearmedizin 2019, 58: 1-9.
View details for DOI 10.1055/a-0830-4453
View details for Web of Science ID 000463226700002
View details for PubMedID 30818412
Comprehensive anatomical and functional imaging in patients with type I neurofibromatosis using simultaneous FDG-PET/MRI
EUROPEAN JOURNAL OF NUCLEAR MEDICINE AND MOLECULAR IMAGING
2019; 46 (3): 776-787
To demonstrate the clinical use of FDG-PET/MRI for monitoring enlargement and metabolism of plexiform neurofibromas (PNF) in patients with neurofibromatosis type 1 (NF1), in whom the development of a malignant peripheral nerve sheath tumor (MPNST) is often a life limiting event.NF1 patients who underwent a simultaneous FDG-PET/MRI examination in our institution from September 2012 to February 2018 were included. Indication was suspicion of malignant transformation of a PNF to MPNST. A maximum of six peripheral nerve lesions per patient were defined as targets. Standardized uptake values (SUV) and apparent diffusion coefficients (ADC) were measured. The presence of target sign and contrast-medium enhancement was visually recorded. Growth rates were estimated comparing prior or follow-up examinations and correlated with FDG uptake and ADC values. The presence of CNS lesions in cerebral T2 weighted images was recorded.In 28 NF1 patients a total number of 83 peripheral nerve tumors, 75 benign PNFs and eight MPNSTs, were selected as target lesions. The SUVs of MPNSTs were significantly higher than the SUVs of PNF (3.84 ± 3.98 [SUVmean MPNSTs] vs. 1.85 ± 1.03 [SUVmean PNF], P < .01). Similarly, lesion SUVmean-to-liver SUVmean ratios significantly differed between MPNSTs and PNF (3.20 ± 2.70 [MPNSTs] vs. 1.23 ± 0.61 [PNF]; P < .01). For differentiation between still benign PNF and MPNSTs, we defined SUVmax ≥ 2.78 as a significant cut-off value. Growth rate of PNF correlated significantly positively with SUVmean (rs = .41; P = .003). MRI parameters like ADCmean (1.87 ± 0.24 × 10-3 mm2/s [PNF] vs. 1.76 ± 0.11 × 10-3 mm2/s [MPNSTs]; P > .05], contrast medium enhancement (P = .50) and target sign (P = .86) did not differ between groups.Simultaneous FDG-PET/MRI is a comprehensive imaging modality for monitoring PNF in NF1 patients. The combined acquisition of both morphologic information in MRI and metabolic information in PET enables the correlation of lesion growth rates with metabolic activity and to define SUV thresholds of significance to identify malignant transformation, which is of utmost clinical significance.
View details for DOI 10.1007/s00259-018-4227-5
View details for Web of Science ID 000457151600024
View details for PubMedID 30535768
Prediction of FDG uptake in Lung Tumors from CT Images Using Generative Adversarial Networks
View details for Web of Science ID 000604567700258
Practice-based evidence for the clinical benefit of PET/CT-results of the first oncologic PET/CT registry in Germany
EUROPEAN JOURNAL OF NUCLEAR MEDICINE AND MOLECULAR IMAGING
2019; 46 (1): 54-64
The purpose of this study was to evaluate the impact of PET/CT on clinical management of cancer patients based on a prospective data registry. The study was developed to inform consultations with public health insurances on PET/CT coverage.We evaluated a prospective patient cohort having a clinically indicated PET/CT at a single German University Center from April 2013 to August 2016. The registry collected questionnaire data from requesting physicians on intended patient management before and after PET/CT. A total of 4,504 patients with 5,939 PET/CT examinations were enrolled in the registry, resulting in evaluable data from 3,724 patients receiving 4,754 scans. The impact of PET/CT on patient management was assessed across 22 tumor types, for different indications (diagnosis, staging, suspected recurrence) and different categories of management including treatment (curative or palliative) and non-treatment (watchful waiting, additional imaging, invasive tests).The most frequent PET/CT indication was tumor staging (59.7%). Melanoma, lung cancer, lymphoma, neuroendocrine tumor and prostate cancer accounted for 70% of cases. Overall, the use of PET/CT resulted in a 37.1% change of clinical management (95% CI, 35.7-38.5), most frequently (30.6%) from an intended non-treatment strategy before PET/CT to active treatment after PET/CT. The frequency of changes ranged from 28.3% for head and neck cancers up to 46.0% for melanomas. The impact of PET/CT was greatest in reducing demands for additional imaging which decreased from 66.1% before PET/CT to 6.1% after PET/CT. Pre-PET/CT planned invasive tests could be avoided in 72.7% of cases. The treatment goal changed after PET/CT in 21.7% of cases, in twice as many cases from curative to palliative therapy than vice versa.The data of this large prospective registry confirm that physicians often change their intended management on the basis of PET/CT by initiating treatment and reducing additional imaging as well as invasive tests. This applies to various cancer types and indications.
View details for DOI 10.1007/s00259-018-4156-3
View details for Web of Science ID 000450924400009
View details for PubMedID 30269155
Value of CT iterative metal artifact reduction in PET/CT-clinical evaluation in 100 patients
BRITISH JOURNAL OF RADIOLOGY
2019; 92 (1096): 20180756
To assess the technical feasibility and diagnostic benefit of CT iterative metal artifact reduction (iMAR) in patients with metal implants undergoing positron emission tomography (PET/CT.PET/CTs of 100 patients with metal implants in different anatomical localization were retrospectively analyzed. CT data were reconstructed with iMAR and without iMAR (noMAR) and used in comparison for PET attenuation correction, generating iMAR-corrected and noMAR PET data. The effect of iMAR on quantitative CT and PET analysis was assessed by measurements of Hounsfield units (HUs) and standard uptake values (SUV) in predefined anatomical structures and pathological lesions in the vicinity of metal implants. Diagnostic confidence for lesion delineation was assessed using a 3-point scale.For artifact-affected structures, mean HU of iMAR corrected CT significantly differed compared to noMAR CT and standard deviations were significantly lower [e.g. M. masseter: 71.01 ± 22.34 HU (iMAR) vs 98.89 ± 92.18 HU (noMAR), p < .01]. SUVs did not significantly differ in artifact-affected structures [e.g. M. masseter: SUVmean 0.96 ± 0.54 (iMAR) vs 0.97 ± 0.55 (noMAR); p > .89] and pathological findings [SUVmean 10.78 (iMAR) vs 10.81 (noMAR); p > .98] between iMAR and noMAR PET. Qualitatively, delineation was significantly improved in iMAR corrected CT for the interpretation of anatomical and pathological structures [e.g. score of pathologic lesions: 2.80 (iMAR) vs 2.31 (noMAR); p < .01].The use of iMAR in PET/CT significantly improves delineation of anatomical and pathological structures in the vicinity of metal implants in CT. PET quantification and PET image quality are not significantly affected by the use of iMAR-based attenuation correction independent of the presence of metal implants.IMAR is a feasible algorithm in PET/CT improving CT image quality in the vicinity of metal implants without affecting PET quantification and can therefore be implemented in the clinical routine.
View details for DOI 10.1259/bjr.20180756
View details for Web of Science ID 000462033100010
View details for PubMedID 30618270
View details for PubMedCentralID PMC6540858
MRI-based assessment and characterization of epicardial and paracardial fat depots in the context of impaired glucose metabolism and subclinical left-ventricular alterations
BRITISH JOURNAL OF RADIOLOGY
2019; 92 (1096): 20180562
To analyze the associations between epicardial and paracardial fat and impaired glucose tolerance as well as left ventricular (LV) alterations.400 subjects underwent 3 T MRI and fat depots were delineated in the four chamber-view of the steady-state free precession cine sequence (repetition time: 29.97 ms; echo time 1.46 ms). LV parameters were also derived from MRI. Oral glucose tolerance tests were performed.Epi- and paracardial fat was derived in 372 (93%) subjects (220 healthy controls, 100 persons with prediabetes, 52 with diabetes). Epi- and paracardial fat increased from normal glucose tolerance (NGT) to prediabetes and diabetes (7.7 vs 9.2 vs 10.3 cm2 and 14.3 vs 20.3 vs 27.4 cm2, respectively; all p < 0.001). However, the association between impaired glucose metabolism and cardiac fat attenuated after adjustment, mainly confounded by visceral adipose tissue (VAT). 93 subjects (27%) had LV impairment, defined as late gadolinium enhancement, ejection fraction < 55% or LV concentricity index > 1.3 g ml-1 . Mean epicardial fat was higher in subjects with LV impairment (11.0 vs 8.1 cm2, p < 0.001). This association remained independent after adjustment for traditional risk factors and VAT [β: 1.13 (0.22; 2.03), p = 0.02].Although epicardial and paracardial fat are increased in prediabetes and diabetes, the association is mostly confounded by VAT. Epicardial fat is independently associated with subclinical LV impairment in subjects without known cardiovascular disease.This study contributes to the picture of epicardial fat as a pathogenic local fat depot that is independently associated with MR-derived markers of left ventricular alterations.
View details for DOI 10.1259/bjr.20180562
View details for Web of Science ID 000462033100005
View details for PubMedID 30633543
View details for PubMedCentralID PMC6540853
ADVERSARIAL INPAINTING OF MEDICAL IMAGE MODALITIES
IEEE. 2019: 3267-3271
View details for Web of Science ID 000482554003098
RETROSPECTIVE CORRECTION OF RIGID AND NON-RIGID MR MOTION ARTIFACTS USING GANS
IEEE. 2019: 1550-1554
View details for Web of Science ID 000485040000328
Unsupervised Medical Image Translation Using Cycle-MedGAN
View details for Web of Science ID 000604567700221
Optimierung der Strahlendosis in der Thorax-CT des Kindes: Einflussfaktoren der Strahlenexposition aus 1695 CT-Untersuchungen in sieben Jahren.
RoFo : Fortschritte auf dem Gebiete der Rontgenstrahlen und der Nuklearmedizin
2018; 190 (12): 1131-1140
PURPOSE: To analyze possible influencing factors on radiation exposure in pediatric chest CT using different approaches for radiation dose optimization and to determine major indicators for dose development.MATERIALS AND METHODS: In this retrospective study at a clinic with maximum care facilities including pediatric radiology, 1695 chest CT examinations in 768 patients (median age: 10years; range: 2 days to 17.9 years) were analyzed. Volume CT dose indices, effective dose, size-specific dose estimate, automatic dose modulation (AEC), and high-pitch protocols (pitch ≥ 3.0) were evaluated by univariate analysis. The image quality of low-dose examinations was compared to higher dose protocols by non-inferiority testing.RESULTS: Median dose-specific values annually decreased by an average of 12 %. High-pitch mode (n = 414) resulted in lower dose parameters (p < 0.001). In unenhanced CT, AEC delivered higher dose values compared to scans with fixed parameters (p < 0.001). In contrast-enhanced CT, the use of AEC yielded a significantly lower radiation dose only in patients older than 16years (p = 0.04). In the age group 6 to 15 years, the values were higher (p < 0.001). The diagnostic image quality of low-dose scans was non-inferior to high-dose scans (2.18 vs. 2.14).CONCLUSION: Radiation dose of chest CT was reduced without loss of image quality in the last decade. High-pitch scanning was an independent factor in this context. Dose reduction by AEC was limited and only relevant for patients over 16 years.KEY POINTS: · The radiation dose of pediatric chest CT was reduced in the last decade.. · High-pitch scanning is an independent factor of dose optimization.. · Dose reduction by AEC is limited and only relevant for older children..CITATION FORMAT: · Esser M, Hess S, Teufel M et al. Radiation Dose Optimization in Pediatric Chest CT: Major Indicators of Dose Exposure in 1695 CT Scans over Seven Years. Fortschr Rontgenstr 2018; 190: 1131 - 1140.
View details for DOI 10.1055/a-0628-7222
View details for PubMedID 30308691
ImFEATbox: a toolbox for extraction and analysis of medical image features
INTERNATIONAL JOURNAL OF COMPUTER ASSISTED RADIOLOGY AND SURGERY
2018; 13 (12): 1881-1893
In medical imaging, the digital post-processing and analysis of acquired images has become an important research field. Topics include various applications of image processing and machine learning aiming to assist radiologists in their diagnostic work. A crucial step in successfully implementing such systems is finding appropriate mathematical descriptions to reflect characteristics of acquired images. Which features are the most meaningful ones strongly depends on the underlying scientific/diagnostic question and the image itself. This makes researching, implementing and testing features time-consuming and cost-intensive. In our work, we aim to address this issue by creating ImFEATbox, a publicly available toolbox to extract and analyze image features for a wide range of applications.To reduce the amount of time spent for choosing the right features, we provide an assortment of feature extraction algorithms which are suitable for a broad variety of medical image processing problems. The toolbox includes both global and local features as well as feature descriptors. While being primarily developed in MATLAB, the majority of our algorithms is also available in Python to enable access to a wider range of researchers.We tested the applicability of ImFEATbox on an FDG-PET/CT data set of 12 patients diagnosed with lung cancer and an MRI data set of 50 patients with prostate lesions. Employing the implemented algorithms in an exemplary manner, we are able to demonstrate its potential for different scientific problems, e.g., show differences between features, indicate redundancies in extracted feature sets by means of a correlation analysis and training a SVM to distinguish between high-risk and low-risk prostate lesions.ImFEATbox provides a variety of feature extraction algorithms suitable for a large number of post-processing and analysis applications in medical imaging. The toolbox is publicly available and can thus be beneficial to a wide range of researchers working on medical image analysis.
View details for DOI 10.1007/s11548-018-1859-7
View details for Web of Science ID 000449775700002
View details for PubMedID 30229363
A machine-learning framework for automatic reference-free quality assessment in MRI
MAGNETIC RESONANCE IMAGING
2018; 53: 134-147
Magnetic resonance (MR) imaging offers a wide variety of imaging techniques. A large amount of data is created per examination which needs to be checked for sufficient quality in order to derive a meaningful diagnosis. This is a manual process and therefore time- and cost-intensive. Any imaging artifacts originating from scanner hardware, signal processing or induced by the patient may reduce the image quality and complicate the diagnosis or any image post-processing. Therefore, the assessment or the ensurance of sufficient image quality in an automated manner is of high interest. Usually no reference image is available or difficult to define. Therefore, classical reference-based approaches are not applicable. Model observers mimicking the human observers (HO) can assist in this task. Thus, we propose a new machine-learning-based reference-free MR image quality assessment framework which is trained on HO-derived labels to assess MR image quality immediately after each acquisition. We include the concept of active learning and present an efficient blinded reading platform to reduce the effort in the HO labeling procedure. Derived image features and the applied classifiers (support-vector-machine, deep neural network) are investigated for a cohort of 250 patients. The MR image quality assessment framework can achieve a high test accuracy of 93.7% for estimating quality classes on a 5-point Likert-scale. The proposed MR image quality assessment framework is able to provide an accurate and efficient quality estimation which can be used as a prospective quality assurance including automatic acquisition adaptation or guided MR scanner operation, and/or as a retrospective quality assessment including support of diagnostic decisions or quality control in cohort studies.
View details for DOI 10.1016/j.mri.2018.07.003
View details for Web of Science ID 000445591900018
View details for PubMedID 30036653
Assessment of image quality of a radiotherapy-specific hardware solution for PET/MRI in head and neck cancer patients
RADIOTHERAPY AND ONCOLOGY
2018; 128 (3): 485-491
Functional PET/MRI has great potential to improve radiotherapy planning (RTP). However, data integration requires imaging with radiotherapy-specific patient positioning. Here, we investigated the feasibility and image quality of radiotherapy-customized PET/MRI in head-and-neck cancer (HNC) patients using a dedicated hardware setup.Ten HNC patients were examined with simultaneous PET/MRI before treatment, with radiotherapy and diagnostic scan setup, respectively. We tested feasibility of radiotherapy-specific patient positioning and compared the image quality between both setups by pairwise image analysis of 18F-FDG-PET, T1/T2-weighted and diffusion-weighted MRI. For image quality assessment, similarity measures including average symmetric surface distance (ASSD) of PET and MR-based tumor contours, MR signal-to-noise ratio (SNR) and mean apparent diffusion coefficient (ADC) value were used.PET/MRI in radiotherapy position was feasible - all patients were successfully examined. ASSD (median/range) of PET and MR contours was 0.6 (0.4-1.2) and 0.9 (0.5-1.3) mm, respectively. For T2-weighted MRI, a reduced SNR of -26.2% (-39.0--11.7) was observed with radiotherapy setup. No significant difference in mean ADC was found.Simultaneous PET/MRI in HNC patients using radiotherapy positioning aids is clinically feasible. Though SNR was reduced, the image quality obtained with a radiotherapy setup meets RTP requirements and the data can thus be used for personalized RTP.
View details for DOI 10.1016/j.radonc.2018.04.018
View details for Web of Science ID 000444505600014
View details for PubMedID 29747873
View details for PubMedCentralID PMC6141811
Voxel-wise correlation of functional imaging parameters in HNSCC patients receiving PET/MRI in an irradiation setup
STRAHLENTHERAPIE UND ONKOLOGIE
2018; 194 (8): 719-726
The purpose of this study was to demonstrate the feasibility of voxel-wise multiparametric characterization of head and neck squamous cell carcinomas (HNSCC) using hybrid multiparametric magnetic resonance imaging and positron emission tomography with [18F]-fluorodesoxyglucose (FDG-PET/MRI) in a radiation treatment planning setup.Ten patients with locally advanced HNSCC were examined with a combined FDG-PET/MRI in an irradiation planning setup. The multiparametric imaging protocol consisted of FDG-PET, T2-weighted transverse short tau inversion recovery sequence (STIR) and diffusion-weighted MRI (DWI). Primary tumours were manually segmented and quantitative imaging parameters were extracted. PET standardized uptake values (SUV) and DWI apparent diffusion coefficients (ADC) were correlated on a voxel-wise level.Images acquired in this specialised radiotherapy planning setup achieved good diagnostic quality. Median tumour volume was 4.9 [1.1-42.1] ml. Mean PET SUV and ADC of the primary tumours were 5 ± 2.5 and 1.2 ± 0.3 10-3 mm2/s, respectively. In voxel-wise correlation between ADC values and corresponding FDG SUV of the tumours, a significant negative correlation was observed (r = -0.31 ± 0.27, p < 0.05).Multiparametric voxel-wise characterization of HNSCC is feasible using combined PET/MRI in a radiation planning setup. This technique may provide novel insights into tumour biology with regard to radiation therapy in the future.
View details for DOI 10.1007/s00066-018-1292-4
View details for Web of Science ID 000439933100003
View details for PubMedID 29564483
- Translational theranostic imaging of lymphoma using radiolabeled alpha CD19-antibodies AMER ASSOC CANCER RESEARCH. 2018
1.5 Tesla MR-Linac Treatment Plans for a Hypoxia-based Dose Escalation in Head and Neck Tumors
SPRINGER HEIDELBERG. 2018: S65
View details for Web of Science ID 000435675600131
Phenotypic Multiorgan Involvement of Subclinical Disease as Quantified by Magnetic Resonance Imaging in Subjects With Prediabetes, Diabetes, and Normal Glucose Tolerance
2018; 53 (6): 357-364
Detailed mechanisms in the pathophysiology of diabetes disease are poorly understood, but structural alterations in various organ systems incur an elevated risk for cardiovascular events and adverse outcome. The aim of this study was to compare multiorgan subclinical disease phenotypes by magnetic resonance (MR) imaging to study differences between subjects with prediabetes, diabetes, and normal controls.Subjects without prior cardiovascular disease were enrolled in a prospective case-control study and underwent multiorgan MR for the assessment of metabolic and arteriosclerotic alterations, including age-related white matter changes, hepatic proton density fat fraction, visceral adipose tissue volume, left ventricular remodeling index, carotid plaque, and late gadolinium enhancement. Magnetic resonance features were summarized in a phenotypic-based score (range, 0-6). Univariate, multivariate correlation, and unsupervised clustering were performed.Among 243 subjects with complete multiorgan MR data sets included in the analysis (55.6 ± 8.9 years, 62% males), 48 were classified as subjects with prediabetes and 38 as subjects with diabetes. The MR phenotypic score was significantly higher in subjects with prediabetes and diabetes as compared with controls (mean score, 3.00 ± 1.04 and 2.69 ± 0.98 vs 1.22 ± 0.98, P < 0.001 respectively), also after adjustment for potential confounders. We identified 2 clusters of MR phenotype patterns associated with glycemic status (P < 0.001), independent of the MR score (cluster II-metabolic specific: odds ratio, 2.49; 95% CI, 1.00-6.17; P = 0.049).Subjects with prediabetes and diabetes have a significantly higher phenotypic-based score with a distinctive multiorgan phenotypic pattern, which may enable improved disease characterization.
View details for DOI 10.1097/RLI.0000000000000451
View details for Web of Science ID 000432188500006
View details for PubMedID 29494349
Plan Comparison Study between a 1.5 T MR-Linac and a Standard-Linac for the definitive Chemoradiotherapy of thoracic Esophageal Cancer
SPRINGER HEIDELBERG. 2018: S151
View details for Web of Science ID 000435675600334
Clinical use of cardiac PET/MRI: current state-of-the-art and potential future applications
JAPANESE JOURNAL OF RADIOLOGY
2018; 36 (5): 313-323
Combined PET/MRI is a novel imaging method integrating the advances of functional and morphological MR imaging with PET applications that include assessment of myocardial viability, perfusion, metabolism of inflammatory tissue and tumors, as well as amyloid deposition imaging. As such, PET/MRI is a promising tool to detect and characterize ischemic and non-ischemic cardiomyopathies. To date, the greatest benefit may be expected for diagnostic evaluation of systemic diseases and cardiac masses that remain unclear in cardiac MRI, as well as for clinical and scientific studies in the setting of ischemic cardiomyopathies. Diagnosis and therapeutic monitoring of cardiac sarcoidosis has the potential of a possible 'killer-application' for combined cardiac PET/MRI. In this article, we review the current evidence and discuss current and potential future applications of cardiac PET/MRI.
View details for DOI 10.1007/s11604-018-0727-2
View details for Web of Science ID 000430656100001
View details for PubMedID 29524169
Automated reference-free detection of motion artifacts in magnetic resonance images
MAGNETIC RESONANCE MATERIALS IN PHYSICS BIOLOGY AND MEDICINE
2018; 31 (2): 243-256
Our objectives were to provide an automated method for spatially resolved detection and quantification of motion artifacts in MR images of the head and abdomen as well as a quality control of the trained architecture.T1-weighted MR images of the head and the upper abdomen were acquired in 16 healthy volunteers under rest and under motion. Images were divided into overlapping patches of different sizes achieving spatial separation. Using these patches as input data, a convolutional neural network (CNN) was trained to derive probability maps for the presence of motion artifacts. A deep visualization offers a human-interpretable quality control of the trained CNN. Results were visually assessed on probability maps and as classification accuracy on a per-patch, per-slice and per-volunteer basis.On visual assessment, a clear difference of probability maps was observed between data sets with and without motion. The overall accuracy of motion detection on a per-patch/per-volunteer basis reached 97%/100% in the head and 75%/100% in the abdomen, respectively.Automated detection of motion artifacts in MRI is feasible with good accuracy in the head and abdomen. The proposed method provides quantification and localization of artifacts as well as a visualization of the learned content. It may be extended to other anatomic areas and used for quality assurance of MR images.
View details for DOI 10.1007/s10334-017-0650-z
View details for Web of Science ID 000428612800002
View details for PubMedID 28932991
Graphical User Interface for Medical Deep Learning - Application to Magnetic Resonance Imaging
IEEE. 2018: 838-847
View details for Web of Science ID 000468383400134
The role of visceral and subcutaneous adipose tissue measurements and their ratio by magnetic resonance imaging in subjects with prediabetes, diabetes and healthy controls from a general population without cardiovascular disease
BRITISH JOURNAL OF RADIOLOGY
2018; 91 (1089): 20170808
To study the relationship of area- and volumetric-based visceral and subcutaneous adipose tissue (VAT and SAT) by MRI and their ratio in subjects with impaired glucose metabolism from the general population.Subjects from a population-based cohort with established prediabetes, diabetes and healthy controls without prior cardiovascular diseases underwent 3 T MRI. VAT and SAT were assessed as total volume and area on a single slice, and their ratio (VAT/SAT) was calculated. Clinical covariates and cardiovascular risk factors, such as hypertension and glycemic state were assessed in standardized fashion. Univariate and adjusted analyses were conducted.Among 384 subjects (age: 56.2 ± 9.2 years, 58.1% male) with complete MRI data available, volumetric and single-slice VAT, SAT and VAT/SAT ratio were strongly correlated (all >r = 0.89). Similarly, VAT/SATvolume ratio was strongly correlated with VATvolume but not with SAT (r = 0.72 and r = -0.21, respectively). Significant higher levels of VAT, SAT and VAT/SAT ratio were found in subjects with impaired glucose metabolism (all p ≤ 0.01). After adjustment for potential cardiovascular confounders, VATvolume and VAT/SATvolume ratio remained significantly higher in subjects with impaired glucose metabolism (VATvolume = 6.9 ± 2.5 l and 3.4 ± 2.3 l; VAT/SATvolume ratio = 0.82 ± 0.34 l and 0.49 ± 0.29 l in patients with diabetes and controls, respectively, all p < 0.02), whereas the association for SATvolume attenuated. Additionally, there was a decreasing effect of glycemic status on VAT/SATvolume ratio with increasing body mass index and waist circumference (p < 0.05).VATvolume and VAT/SATvolume ratio are associated with impaired glucose metabolism, independent of cardiovascular risk factors or MRI-based quantification technique, with a decreasing effect of VAT/SATvolume ratio in obese subjects. Advances in knowledge: Quantification of VATvolume and VAT/SATvolume ratio by MRI represents a reproducable biomarker associated with cardiometabolic risk factors in subjects with impaired glucose metabolism, while the association of VAT/SATvolume ratio with glycemic state is attenuated in obese subjects.
View details for DOI 10.1259/bjr.20170808
View details for Web of Science ID 000443131900013
View details for PubMedID 29388794
View details for PubMedCentralID PMC6223151
AUTOMATIC MOTION ARTIFACT DETECTION FOR WHOLE-BODY MAGNETIC RESONANCE IMAGING
IEEE. 2018: 995-999
View details for Web of Science ID 000446384601038
AUTOMATED DETECTION OF HIGH FDG UPTAKE REGIONS IN CT IMAGES
IEEE. 2018: 1065-1069
View details for Web of Science ID 000446384601051
SEMANTIC ORGAN SEGMENTATION IN 3D WHOLE-BODY MR IMAGES
IEEE. 2018: 3498-3502
View details for Web of Science ID 000455181503123
MR Image Reconstruction Using a Combination of Compressed Sensing and Partial Fourier Acquisition: ESPReSSo
IEEE TRANSACTIONS ON MEDICAL IMAGING
2016; 35 (11): 2447-2458
A Cartesian subsampling scheme is proposed incorporating the idea of PF acquisition and variable-density Poisson Disc (vdPD) subsampling by redistributing the sampling space onto a smaller region aiming to increase k-space sampling density for a given acceleration factor. Especially the normally sparse sampled high-frequency components benefit from this sampling redistribution, leading to improved edge delineation. The prospective subsampled and compacted k-space can be reconstructed by a seamless combination of a CS-algorithm with a Hermitian symmetry constraint accounting for the missing part of the k-space. This subsampling and reconstruction scheme is called Compressed Sensing Partial Subsampling (ESPReSSo) and was tested on in-vivo abdominal MRI datasets. Different reconstruction methods and regularizations are investigated and analyzed via global (intensity-based) and local (region-of-interest and line evaluation) image metrics, to conclude a clinical feasible setup. Results substantiate that ESPReSSo can provide improved edge delineation and regional homogeneity for multidimensional and multi-coil MRI datasets and is therefore useful in applications depending on well-defined tissue boundaries, such as image registration and segmentation or detection of small lesions in clinical diagnostics.
View details for DOI 10.1109/TMI.2016.2577642
View details for Web of Science ID 000388503400008
View details for PubMedID 27295659
Evaluation of MR imaging properties of a dedicated RT positioning solution for combined PET/MR imaging
ELSEVIER IRELAND LTD. 2015: S222
View details for Web of Science ID 000688178300453
Hypoxia imaging and functional MR using combined FMISO PET/MRI in head and neck cancer (HNC)
ELSEVIER IRELAND LTD. 2014: S53
View details for Web of Science ID 000688272900129
Evaluation of a dedicated radiotherapy positioning solution for combined PET/MR imaging
ELSEVIER IRELAND LTD. 2014: S33
View details for Web of Science ID 000688272900079