Скачать или смотреть EEG Paradox TraceRoute

Working on something for fun and exploration called EEG Traceroute. It borrows the concept from internet traceroute and applies it to EEG connectivity you pick a source electrode and it traces the signal path hop by hop through the network, using Granger causality for direction and strength at each step.

Three routing modes: Greedy follows the strongest connection at each hop. Dijkstra finds the lowest-cost path to the network's primary sink. Cascade expands outward from the source to show how the network gets recruited wave by wave.
Each frequency band (delta through gamma) gets its own connectivity graph, so you can see how routing changes across bands. There's an interactive 3D view where you can switch the source, band, and policy.

Standard EEG connectivity gives you pairwise relationships "A is connected to B." This just chains those relationships together and asks where the path actually goes. Every edge in the graph comes from Granger causality with FDR correction, so it's working from statistically tested connections, not arbitrary ones.