Data-driven cortical clustering to provide a family of plausible solutions to M/EEG inverse problem
This addresses the inverse problem in neuroimaging for researchers, offering a family of plausible solutions instead of a single one, though it is incremental as it builds on existing constraints.
The paper tackled the ill-posed M/EEG inverse problem by assuming brain activity is a connected cortical region, showing that even with one active region, multiple configurations can explain the data, and proposed a method to find several good candidate regions with similar accuracy.
The M/EEG inverse problem is ill-posed. Thus additional hypotheses are needed to constrain the solution space. In this work, we consider that brain activity which generates an M/EEG signal is a connected cortical region. We study the case when only one region is active at once. We show that even in this simple case several configurations can explain the data. As opposed to methods based on convex optimization which are forced to select one possible solution, we propose an approach which is able to find several "good" candidates - regions which are different in term of their sizes and/or positions but fit the data with similar accuracy.