Size and shape calibration gave a 0.002 lower mean voltage agreement at a 50 percent pair budget in eight SEEG sessions

A brain-computer interface needs a reliable way to measure the response to electrical stimulation. This study tested two five-pulse calibration rules in public SEEG data from eight epilepsy patients. One rule ranked waveform shape. The other rule ranked waveform size and shape. Each rule selected the same number of pairs at each site. The final five pulses measured later voltage agreement, without use of calibration pulses in that outcome. The study used 884 stimulation-to-readout pairs at 84 sites. Mean later agreement was 0.594 for size and shape calibration and 0.596 for shape calibration. The paired change was −0.00213, with a patient bootstrap 95 percent interval from −0.00861 to +0.00402. The exact two-sided sign-flip p value was 0.6406. Only 14 of 463 pair choices changed. The data do not show a benefit from the new rank rule in the tested sessions. The package contains exact trial products, five figures, code, source hashes, a fresh source check, and an offline reproduction test. The result measures centered voltage within a session. It does not test useful BCI control, neural origin, or long-term stability.

stereoelectroencephalography · electrical stimulation · waveform agreement · brain-computer interface · concordance correlation · calibration