Pericentral SEEG pairs showed a small periodic beta excess over other frontoparietal pairs in 90 epilepsy patients

A preregistered confirmatory analysis tested whether periodic beta power was higher at pericentral than at other frontoparietal stereo-electroencephalography (SEEG) bipolar pairs in the same patient. We analyzed archived 0.5 Hz Welch spectra from 60 s segments beginning at the first WORD event in RAM free-recall tasks. The sample included 90 patients, 1004 pericentral pairs, 4679 comparison pairs, and 8 sites. Periodic beta, defined as the maximum 13–30 Hz residual above a 2–40 Hz log-log aperiodic line, was higher pericentrally by 0.050 log10 units (95% CI 0.020–0.082; 55/90 patients positive; Holm p = 0.0015). Relative beta was higher by 0.095 (95% CI 0.064–0.128; Holm p = 0.0001). A flank-fit periodic beta measure also increased (+0.077, p = 0.0021), while the aperiodic exponent decreased (−0.170, p = 0.00005). Peak frequency did not differ (−0.03 Hz, p = 0.93; approximately 18.4 Hz in both groups). Effects remained significant when each site was excluded in turn. Against frontoparietal pairs within 30 mm, effects were smaller and at the significance boundary: periodic beta +0.035 (p = 0.053), relative beta +0.039 (p = 0.057), and aperiodic exponent −0.087 (p = 0.041; 82 patients). These results support a small periodic beta excess, while the larger relative-beta difference also reflects a flatter pericentral aperiodic spectrum. The distance sensitivity is consistent with a spatial gradient and does not establish a sharp boundary. Data came from patients with epilepsy during a memory task; labels were atlas labels of bipolar pairs, there was one segment per patient, and pairs in the seizure onset zone were not excluded.

stereo-electroencephalography · beta oscillations · pericentral cortex · aperiodic spectrum · memory · epilepsy · confirmatory analysis