BCIwiki (bciwiki.com) — According to Donders Centre for Neuroscience, Department of Neurophysics, a study on improved neural traveling-wave detection and spatiotemporal spectral analysis was uploaded to bioRxiv on 2026-09-16. The preprint has not been peer reviewed. It introduces the Traveling Wave Index (TWINDEX) framework and applies it to marmoset electrocorticography (ECoG) and human EEG data.
Neural traveling waves are patterns of neural activity that propagate through space over time. TWINDEX analyzes temporal frequency, spatial frequency, and propagation direction, and accommodates irregular sensor layouts. It quantifies wave strength as the reduction in circular phase variance produced by a candidate planar wave, assessing how well that propagation pattern accounts for phase differences across locations. The authors report that this normalization behaves robustly at both low and high spatial frequencies and suppresses coherent in-phase activity.
The study uses directional moments to distinguish planar from standing waves and derives analytical links between TWINDEX, parametric planar-wave fitting, and distance-phase correlation. The researchers also introduce projected distance-phase correlation (ProDPC) for sensitive detection of planar waves in individual trials.
In large-scale marmoset ECoG and human EEG data, the methods identified alpha and low-beta traveling waves concentrated in specific spatial and temporal frequency ranges, with speeds the authors considered physiologically plausible. Across trials and epochs, waves occurred in oppositely directed propagation modes. The authors interpret this as evidence that alpha and beta activity is associated with both feedforward- and feedback-directed large-scale dynamics, rather than proof of a specific causal role for either propagation direction.