Enhancing Precision in Cardiac Segmentation for Magnetic Resonance-Guided Radiation Therapy Through Deep Learning.

PURPOSE: Cardiac substructure dose metrics are more strongly linked to late cardiac morbidities than to whole-heart metrics. Magnetic resonance (MR)-guided radiation therapy (MRgRT) enables substructure visualization during daily localization, allowing potential for enhanced cardiac sparing. We extend a publicly available state-of-the-art deep learning framework, "No New" U-Net, to incorporate self-distillation (nnU-Net.wSD) for substructure segmentation for MRgRT. METHODS AND MATERIALS: Eighteen (institute A) patients who underwent thoracic or abdominal radiation therapy on a 0.35 T MR-guided linear accelerator were retrospectively evaluated. On each image, 1 of 2 radiation oncologists delineated reference contours of 12 cardiac substructures (chambers, great vessels, and coronary arteries) used to train (n = 10), validate (n = 3), and test (n = 5) nnU-Net.wSD by leveraging a teacher-student network and comparing it to standard 3-dimensional U-Net. The impact of using simulation data or including 3 to 4 daily images for augmentation during training was evaluated for nnU-Net.wSD. Geometric metrics (Dice similarity coefficient, mean distance to agreement, and 95% Hausdorff distance), visual inspection, and clinical dose-volume histograms were evaluated. To determine generalizability, institute A's model was tested on an unlabeled data set from institute B (n = 22) and evaluated via consensus scoring and volume comparisons. RESULTS: nnU-Net.wSD yielded a Dice similarity coefficient (reported mean ± SD) of 0.65 ± 0.25 across the 12 substructures (chambers, 0.85 ± 0.05; great vessels, 0.67 ± 0.19; and coronary arteries, 0.33 ± 0.16; mean distance to agreement, <3 mm; mean 95% Hausdorff distance, <9 mm) while outperforming the 3-dimensional U-Net (0.583 ± 0.28; P <.01). Leveraging fractionated data for augmentation improved over a single MR simulation time point (0.579 ± 0.29; P <.01). Predicted contours yielded dose-volume histograms that closely matched those of the clinical treatment plans where mean and maximum (ie, dose to 0.03 cc) doses deviated by 0.32 ± 0.5 Gy and 1.42 ± 2.6 Gy, respectively. There were no statistically significant differences between institute A and B volumes (P >.05) for 11 of 12 substructures, with larger volumes requiring minor changes and coronary arteries exhibiting more variability. CONCLUSIONS: This work is a critical step toward rapid and reliable cardiac substructure segmentation to improve cardiac sparing in low-field MRgRT.

Spatial-Temporal Synchronous Graph Transformer network (STSGT) for COVID-19 forecasting.

COVID-19 has become a matter of serious concern over the last few years. It has adversely affected numerous people around the globe and has led to the loss of billions of dollars of business capital. In this paper, we propose a novel Spatial-Temporal Synchronous Graph Transformer network (STSGT) to capture the complex spatial and temporal dependency of the COVID-19 time series data and forecast the future status of an evolving pandemic. The layers of STSGT combine the graph convolution network (GCN) with the self-attention mechanism of transformers on a synchronous spatial-temporal graph to capture the dynamically changing pattern of the COVID time series. The spatial-temporal synchronous graph simultaneously captures the spatial and temporal dependencies between the vertices of the graph at a given and subsequent time-steps, which helps capture the heterogeneity in the time series and improve the forecasting accuracy. Our extensive experiments on two publicly available real-world COVID-19 time series datasets demonstrate that STSGT significantly outperforms state-of-the-art algorithms that were designed for spatial-temporal forecasting tasks. Specifically, on average over a 12-day horizon, we observe a potential improvement of 12.19% and 3.42% in Mean Absolute Error (MAE) over the next best algorithm while forecasting the daily infected and death cases respectively for the 50 states of US and Washington, D.C. Additionally, STSGT also outperformed others when forecasting the daily infected cases at the state level, e.g., for all the counties in the State of Michigan. The code and models are publicly available at https://github.com/soumbane/STSGT.

Deep reasoning neural network analysis to predict language deficits from psychometry-driven DWI connectome of young children with persistent language concerns.

This study investigated whether current state-of-the-art deep reasoning network analysis on psychometry-driven diffusion tractography connectome can accurately predict expressive and receptive language scores in a cohort of young children with persistent language concerns (n = 31, age: 4.25 ± 2.38 years). A dilated convolutional neural network combined with a relational network (dilated CNN + RN) was trained to reason the nonlinear relationship between "dilated CNN features of language network" and "clinically acquired language score". Three-fold cross-validation was then used to compare the Pearson correlation and mean absolute error (MAE) between dilated CNN + RN-predicted and actual language scores. The dilated CNN + RN outperformed other methods providing the most significant correlation between predicted and actual scores (i.e., Pearson's R/p-value: 1.00/<.001 and .99/<.001 for expressive and receptive language scores, respectively) and yielding MAE: 0.28 and 0.28 for the same scores. The strength of the relationship suggests elevated probability in the prediction of both expressive and receptive language scores (i.e., 1.00 and 1.00, respectively). Specifically, sparse connectivity not only within the right precentral gyrus but also involving the right caudate had the strongest relationship between deficit in both the expressive and receptive language domains. Subsequent subgroup analyses inferred that the effectiveness of the dilated CNN + RN-based prediction of language score(s) was independent of time interval (between MRI and language assessment) and age of MRI, suggesting that the dilated CNN + RN using psychometry-driven diffusion tractography connectome may be useful for prediction of the presence of language disorder, and possibly provide a better understanding of the neurological mechanisms of language deficits in young children.

Deep Relational Reasoning for the Prediction of Language Impairment and Postoperative Seizure Outcome Using Preoperative DWI Connectome Data of Children With Focal Epilepsy.

Prolonged seizures in children with focal epilepsy (FE) may impair language functions and often reoccur after surgical intervention. This study is aimed at developing a novel deep relational reasoning network to investigate whether conventional diffusion-weighted imaging connectome analysis can be improved when predicting expressive and receptive scores of preoperative language impairments and classifying postoperative seizure outcomes (seizure freedom or recurrence) in individual FE children. To deeply reason the dependencies of axonal connections that are sparsely distributed in the whole brain, this study proposes the "dilated CNN + RN", a dilated convolutional neural network (CNN) combined with a relation network (RN). The performance of the dilated CNN + RN was evaluated using whole brain connectome data from 51 FE children. It was found that when compared with other state-of-the-art algorithms, the dilated CNN + RN led to an average improvement of 90.2% and 97.3% in predicting expressive and receptive language scores, and 2.2% and 4% improvement in classifying seizure freedom and seizure recurrence, respectively. These improvements were independent of the prefixed connectome densities. Also, the dilated CNN + RN could provide an explainable artificial intelligence (AI) model by computing gradient-based regression/classification activation maps. This mapping analysis revealed left superior-medial frontal cortex, bilateral hippocampi, and cerebellum as crucial hubs, facilitating important connections that were most predictive of language function and seizure refractoriness after surgery.

Simulations of a Multi-Pinhole SPECT Collimator for Clinical Dopamine Transporter (DAT) Imaging.

SPECT imaging of the dopamine transporter (DAT) is used for diagnosis and monitoring progression of Parkinson's Disease (PD), and differentiation of PD from other neurological disorders. The diagnosis is based on the DAT binding in the caudate and putamen structures in the striatum. We previously proposed a relatively inexpensive method to improve the detection and quantification of these structures for dual-head SPECT by replacing one of the fan-beam collimators with a specially designed multi-pinhole (MPH) collimator. In this work, we developed a realistic model of the proposed MPH system using the GATE simulation package and verified the geometry with an analytic simulator. Point source projections from these simulations closely matched confirming the accuracy of the pinhole geometries. The reconstruction of a hot-rod phantom showed that 4.8 mm resolution is achievable. The reconstructions of the XCAT brain phantom showed clear separation of the putamen and caudate, which is expected to improve the quantification of DAT imaging and PD diagnosis. Using this GATE model, point spread functions modeling physical factors will be generated for use in reconstruction. Also, further improvements in geometry are being investigated to increase the sensitivity of this base system while maintaining a target spatial resolution of 4.5-5 mm.