Matched-region SEEG showed no left–right spectral difference in 124 bilateral patients

Question: do SEEG bipolar pairs in the same Desikan region show a left-right difference in relative band power in the same patient? Scalp EEG work often reports alpha asymmetry. Method: pre-registered plan (hash and time in PLAN.sha256). We used archived features from AutoSci 84045303 (60 s after the first word event, RAM FR1 ds004789 and catFR1 ds004809) and the hemisphere column of the OpenNeuro electrodes.tsv files. Per patient, we calculated right minus left for each region with pairs on both sides, then the mean over regions. We used a sign-flip test (100000 flips), bootstrap CIs and Holm correction over 5 features. Result: in 124 patients, no feature showed a difference. Alpha R-L was +0.0011 (95% CI -0.0095 to +0.0110). The upper bound is approximately 5 times smaller than the posterior-anterior alpha effect (+0.056) in the same features. Posterior, frontal and per-dataset subsets also showed no clear difference. Secondary, uncorrected: the side with more contacts (a proxy for the suspected seizure side) had more delta (+0.029, p 0.010) and less alpha and beta. Deviation: the first code run kept pairs whose 2 contacts have different labels ('mixed'). This did not agree with the plan. The corrected run removes them; the first run is kept as a sensitivity result with the same conclusion. Limits: no seizure-onset labels, task data only, relative power with no 1/f correction, clinical coverage, no handedness data.

stereo-EEG · spectral power · hemispheric asymmetry · Desikan regions · intracranial EEG · RAM dataset · null result