Improved detection and spatiotemporal spectral analysis of neural traveling waves

Traveling waves (TWs) are key neural dynamics, but current detection methods face challenges due to sensor constraints, spatial-frequency limits, signal intensity, and confusion between moving and stationary wave patterns. To address these issues, researchers have developed the Traveling Wave Index (TWINDEX), a three-dimensional spatiotemporal analysis framework that examines TWs across frequency, spatial frequency, and direction of propagation.

TWINDEX can be applied to non-standard sensor arrangements and measures wave intensity by evaluating the decrease in circular phase variance produced by a hypothetical planar wave. This normalization ensures consistent performance at low and high spatial frequencies and minimizes the impact of coherent in-phase activity. To further distinguish between planar and standing waves, directional moments are used.

By connecting TWINDEX to parametric planar-wave fitting and distance-phase correlation, a method called projected distance-phase correlation (ProDPC) is introduced for reliable single-trial planar-wave detection. When applied to extensive marmoset ECoG and human EEG data, TWINDEX and ProDPC reveal alpha/low-beta TWs localized in both spatial and temporal frequency domains.

These waves demonstrate physiologically plausible propagation speeds and exhibit propagation in opposing directions, highlighting that alpha/beta activity is linked to both feedforward and feedback-driven large-scale neural dynamics.