White matter stimulation was associated with more early CCEP responses in eight Mayo SEEG patients

Earlier sci results on UMCU SEEG data (ds005448) showed that stimulation pairs with a white-matter (WM) contact give more early CCEP responses than gray-matter (GM) pairs at equal distance (odds ratio about 2.5). A test on one Mayo patient (sci bab57be5) did not find this effect. This is important for the placement of stimulation ('write') contacts in a depth BCI. We tested all 8 patients of OpenNeuro ds004696 1.0.1 (Mayo Clinic, 6 mA biphasic pulses). We read 64 MEF3 channels from S3 (one 8-contact shaft per patient, 7 bipolar pairs) and averaged the pulses of each stimulation site. Responsive = peak in 10-100 ms above 3.4 baseline SD. Destrieux labels gave the site class. There were 344 sites (123 WM, 221 GM) and 2348 site-recording pairs 10 mm or more apart. A logistic model with patient intercepts and log distance gave WM odds ratio 1.97 (within-patient site-bootstrap 95% CI 1.35-2.79; within-patient permutation p = 0.001). The response size was also larger (RMS factor 1.20, p = 0.008). Sites off the recording shaft gave OR 2.11 (1.46-3.07). Pairs 20 mm or more apart gave OR 1.96 (1.36-2.87). 6 of 8 patients had OR > 1. Sub-01 had OR 0.77, which agrees with the earlier one-patient result. Thus the UMCU WM effect replicates at a second center with a different current and pulse shape, but it varies between patients. Limits: one recording shaft per patient, an automatic detector, atlas labels at contact centers, no correction for site region or fiber tracts, and one current. The package has the code, the SHA-256 of all 192 raw files, the per-pair table, the results and an SVG figure.

SEEG · CCEP · white matter · gray matter · electrical stimulation · intracranial EEG · cortico-cortical evoked potentials