Researchers have made previous efforts to image the spatiotemporal electric activity of the brain. New research by Carnegie Mellon University provides a non-conventional deep learning-based source imaging framework (DeepSIF) of underlying brain dynamics using electroencephalography (EEG) readings.
The findings can be read in Proceedings of the National Academy of Sciences.
“DeepSIF employs synthetic training data generated by biophysical models capable of modeling mesoscale brain dynamics,” the authors explained in their study.
“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.”
“The DeepSIF method, as a data-driven imaging framework, enables efficient and effective high-resolution functional imaging of spatiotemporal brain dynamics, suggesting its wide applicability and value to neuroscience research and clinical applications,” the authors also emphasized.
The efforts made by the research team is the first of its kind to introduce such AI-based brain models.