Fig. 1: The concept of the proposed spatiotemporal source imaging approach. | Nature Communications

Fig. 1: The concept of the proposed spatiotemporal source imaging approach.

From: Noninvasive electromagnetic source imaging of spatiotemporally distributed epileptogenic brain sources

Fig. 1: The concept of the proposed spatiotemporal source imaging approach.The alternative text for this image may have been generated using AI.

Brain networks are modeled as focally extended sources that vary over time. The net-effect of these dynamics are recorded in EEG/MEG. Blind source separation (BSS) techniques applied to these measurements can delineate these underlying dynamics and serve as a temporal prior in the imaging algorithm. Spatial constraints that enforce the edge sparsity, i.e. clear distinction of activity and background noise, can be ensured by applying an iterative reweighting scheme in a data-driven manner to guarantee focally extended sources. Combining these data-driven priors into our imaging module, we can estimate underlying brain networks; the nodes and internodal connectivity (links) of these networks. Nodes are spatially extended regions in the brain and not focal points. Our spatiotemporal source imaging approach considers the functional segregation, i.e. spatially coherent regions in the brain specialized for specific functions, and functional integration, i.e. inter-regional communication and connectivity, of different brain regions.

Back to article page