Theta and alpha phase lags show no anterior-posterior direction in HUP SEEG

This analysis tested whether interictal theta and alpha phase lags between distant sites in stereo-electroencephalography (SEEG) show a consistent anterior-posterior (A-P) direction. It used run-01 recordings from the HUP epilepsy dataset at https://s3.amazonaws.com/openneuro.org/ds004100. Thirty-six of 37 identified subjects were processed; 30 met the inclusion rule. For each patient, imaginary coherency between bipolar channels on different shafts was modeled against A-P, superior-inferior, and medial-lateral separation. The theta A-P coefficient was +0.0001 per 10 mm (95% CI -0.0004 to +0.0008; 10/30 positive; Holm p = 0.67); the alpha coefficient was +0.0002 per 10 mm (95% CI -0.00004 to +0.0006; 16/30 positive; Holm p = 0.27). The findings support neither posterior-to-anterior nor anterior-to-posterior direction for the measured whole-brain lag trend. Secondary analyses found no directional effect in delta, theta, alpha, or beta across the modeled axes (all p >= 0.07, uncorrected); strong distant pairs showed posterior lead in 0.49-0.51 of cases across bands. A within-shaft test found the deeper bipolar pair led the adjacent superficial pair in all bands (mean imaginary coherency +0.05 to +0.09; 22-25 of 30 positive; p = 0.0001-0.0007). A post-hoc phase-angle increase with frequency corresponded to an approximately constant 3-7 ms lag, which may reflect sequential channel sampling, bipolar arithmetic with unequal contact impedance, or conduction; these causes cannot be separated, and this finding is not interpreted as a travelling wave. Limits include unlabelled wake/sleep state, epilepsy patients, clinical coverage, fsaverage coordinates, sparse A-P sampling, and the distinction between a whole-brain linear trend and local waves with mixed directions.

stereo-electroencephalography · SEEG · imaginary coherency · phase lag · theta · alpha · anterior-posterior · epilepsy