Interactive hands-on · Functional Connectivity with fNIRS
This is the main page for the hands-on sessions for the mini-course "Advanced Topics in Functional Connectivity with fNIRS: from isolated to integrated approaches," delivered at the fNIRS 2026 conference in Macau, China. The sessions follow the structure of part 2 of the course, which is focused on graph theory and its application to resting-state fNIRS functional connectivity. The hands-on sessions are designed to be interactive and run entirely in your web browser, allowing you to explore real data and understand the properties of the whole system through a series of short exercises. Everything runs in your browser on real data: move a slider, and the graph and its numbers update.
Significance and autocorrelation, negative correlations, absolute and proportional thresholds, binary or weighted, and the threshold as a control parameter.
Open → Hands-on 2Clustering, path length and efficiency against random graphs, across densities and chromophores; the Watts–Strogatz model; coarse-graining.
Open → Hands-on 3Degree, strength, betweenness and eigenvector centrality; Louvain vs spectral modules; provincial and connector hubs; random failure vs targeted attack.
Open → Hands-on 4Distance between channels and how to correct for it, and how channel pruning changes the graph.
Open → Free explorationEvery control on one screen, including signal, nodes, negatives, threshold, weights, centrality, modules and resilience.
Open →One healthy adult, 8 minutes at rest with eyes open, whole-head continuous-wave fNIRS. The connectivity matrix comes from the Part 1 pipeline: quality control (SCI/PSP), TDDR and wavelet motion correction, MBLL, band-pass 0.009–0.08 Hz, short-channel regression, and Pearson correlation for HbO, HbR and HbT. Channel regions are approximate (closest AAL region on a template head).
HandsOn1…4_*.m) if you want to see how each number is computed (not availabe for this course, sorry).