Interictal SEEG band amplitude showed no corrected locking to ECG-derived breathing phase in 17 patients

We tested whether theta (4–8 Hz), beta (13–30 Hz), and high-frequency (70–110 Hz) interictal stereoelectroencephalography (SEEG) amplitude locked to breathing in patients with epilepsy. Because the public SEEG dataset had no respiration channel, we estimated breathing phase from ECG R-peak amplitude. We analyzed the first 300 s from 17 patients and 1,322 bipolar pairs in OpenNeuro ds004100. We excluded samples from 0.10 s before to 0.35 s after each R peak and compared mean vector length with 200 circular-shift surrogates. In the 9 patients with valid ECG-derived respiration, the median fraction of significant high-frequency pairs was 0.074; it was not above chance after Holm correction (one-sided Wilcoxon p = 0.10, Holm p = 0.30). Theta and beta showed no evidence of locking. Across all 17 patients, high-frequency activity showed a weak trend (median fraction 0.087; p = 0.036 uncorrected, Holm p = 0.11), and real phase exceeded the pseudo-breathing control (p = 0.03). Preferred phase was not consistent across patients. These results do not support the primary hypothesis. ECG-derived phase is indirect, recordings were short, and processing stopped after 21 of 37 listed subjects. The analysis used NumPy-only FFT filters because SciPy was unavailable; ECG-derived respiration interpolation was linear.

stereoelectroencephalography · breathing · ECG-derived respiration · phase-amplitude coupling · epilepsy · high-frequency activity · negative result