In five patients, high-band SEEG stimulation artifact decreased with distance but often exceeded baseline at 60–90 mm

We measured stimulation-artifact spread in bipolar stereo-electroencephalography (SEEG) channels during 5 s, 50 Hz biphasic trains using 10 recordings from five patients in OpenNeuro ds006519 v1.0.0. For each channel, we calculated the root-mean-square (RMS) ratio between the detected train and the 5 s prestimulation interval in 250–1000 Hz and 2–15 Hz bands, and related this ratio to distance from the stimulated pair. In the high band, the median ratio was 18.0 within 20 mm and 3.4 at 60 mm or more. The median log–log slope was −1.37 (five of five subjects had negative slopes; exact sign-test p = 0.0625). The fitted ratio reached 2 at a median distance of 93 mm (range 57–432 mm). In the low band, RMS did not increase at any distance and decreased to 0.63 within 20 mm; the data cannot distinguish neural suppression from a recording effect. The analysis included five subjects and 10 of 35 recordings; the other 25 were not analyzed because of a download time limit. Limits include 50 Hz trains only, mixed currents of 0.75–3 mA with current effects not modeled, same-shaft bipolar montage, native-space coordinates, estimated positions for 13 of 20 stimulated contacts, and RMS ratios that include neural responses as well as artifact. The high band is above the 50 Hz fundamental but contains its harmonics. Code, per-channel results, figure, and the S3 key list are included.

stereo-electroencephalography · SEEG · electrical stimulation · stimulation artifact · bipolar recording · signal contamination · artifact spread · intracranial EEG