Deep learning-based Source Imaging Improves Spatiotemporal Imaging of Epileptic Sources

 Many efforts have been made to image the spatiotemporal electrical activity of the brain, with the purpose of mapping its function and dysfunction, as well as aiding the management of brain disorders. Here, I develop a novel deep learning-based source imaging framework (DeepSIF), that provides robust and precise spatiotemporal estimates of underlying brain dynamics from non-invasive high-density electroencephalography (EEG) and magnetoencephalography (MEG) recordings. DeepSIF employs synthetic training data generated by biophysical models capable of modeling mesoscale brain dynamics. The rich characteristics of underlying brain sources are embedded in the realistic training data and implicitly learned by DeepSIF networks, avoiding complications associated with explicitly formulating and tuning priors in an optimization problem, as often is the case, in conventional source imaging approaches. 

I first demonstrate the necessity of using a realistic source model, such as a brain network model consisting of interconnected neural mass models, for the training data. Then, deep neural networks are trained and evaluated by (1) series of comprehensive numerical experiments, (2) imaging sensory and cognitive brain responses in a total of 20 healthy subjects from three public datasets, and (3) rigorously validating DeepSIF’s capability in identifying epileptogenic regions in a total cohort of 64 drug-resistant epilepsy patients by comparing DeepSIF results with invasive measurements and surgical resection outcomes. Interictal spike from EEG or MEG recordings, as well as ictal recordings from EEG recordings from three different clinical centers are used as the input for the trained DeepSIF networks. 

DeepSIF demonstrates robust and excellent performance, producing results that are concordant with common neuroscience knowledge about sensory and cognitive information processing, as well as clinical findings about the location and extent of the epileptogenic tissue, outperforming conventional source imaging methods. The DeepSIF method, as a novel data-driven imaging framework, demonstrates its versatility at adapting to different tasks as a general ESI framework. It enables efficient and effective high-resolution functional imaging of spatiotemporal brain dynamics, suggesting its wide applicability and value to neuroscience research and clinical applications.