In three SEEG patients, median recovery at 40 mm or more was 3.4 ms bipolar and 9.3 ms monopolar

We measured how long the stimulation artefact stays in SEEG channels after one 1 ms, 1-2 mA pulse. We used archived epochs (-40 to +80 ms) of 3 patients (STREEF06-08, 60 pulses each, 30 sites each) from OpenNeuro ds005448. For each channel and pulse, the recovery time T is the start of the first 5 ms run where the absolute first difference is less than 5 times the robust baseline noise. We compared monopolar signals (as stored) with adjacent bipolar signals on the same shaft, as a function of the distance to the nearest stimulated contact. In 21,802 channel-pulses, monopolar T had a median of approximately 9-10 ms at all distances up to 100 mm. Adjacent bipolar T decreased to a median of 3.4 ms at 40 mm or more. At 10 mm or more, a blanking of 36 ms (monopolar) or 26 ms (bipolar) covered 99% of channel-pulses (range over patients: 17-45 ms and 16-33 ms). Within 10 mm, the 99th percentile was more than 80 ms (censored). Spearman rho between T and distance was -0.34 (monopolar) and -0.52 (bipolar). With no pulse, the same rule gave T > 5 ms in only 1.2-1.8% of cases. Thresholds of 3 and 8 sigma changed the values but not the monopolar-vs-bipolar difference at 40 mm or more. Limits: 3 patients, one amplifier setting, 80 ms window, and the rule cannot separate artefact from early neural responses near the site. The package has code, per-pulse results, a figure and input checksums.

stereo-EEG · SEEG · electrical stimulation · stimulation artefact · artefact recovery · bipolar reference · monopolar reference · blanking period · closed-loop brain-computer interface · signal processing