Mean SEEG residual agreement was 0.786 before and 0.777 after pulses in five patients at 100 to 500 ms

We measured shared trial voltage in public SEEG data from five patients in OpenNeuro ds004457 version 1.0.2. We removed each pulse's window mean and each pulse group's mean response. The score multiplied correlations between receiving contacts from two groups of five pulses. We compared matched 100 to 500 ms windows before and after stimulation. The analysis used 120 blocks, 1,200 pulses, and 19 contacts. The equal patient mean score was 0.786033 before and 0.777277 after pulses. The change was −0.008756. The patient bootstrap 95 percent interval was [−0.024505, +0.006993]. The exact two-sided patient sign test gave p = 0.375. Population inference requires independent patients and the stated sampling or null sign assumptions. Five patients give a smallest two-sided sign p value of 0.0625. Excluding pulses with status notes changed the mean contrast to +0.015362. A bipolar check used only three patients and gave a change of +0.004935. The observed change thus depended on selection and processing. The data do not establish equal population means or neural connections. Public raw files reproduced every input array. Independent formulas, invariance checks, and simulations passed. The package preserves data, code, source pins, matrices, selection records, inference draws, figures, and reproduction instructions. This is a new analysis of previously used patients, not an independent patient sample.

stereoelectroencephalography · SEEG · trial covariance · residual correlation · electrical stimulation · ventral temporal cortex · patient bootstrap · signal processing