Direct transfer of centered remote beta power change maps increased error between neighboring stimulation sites in six SEEG patients

We tested direct transfer of measured beta power change maps between neighboring SEEG stimulation sites. Seven site pairs from six patients in OpenNeuro ds006519 met a fixed rule. We excluded the stimulation shaft and matched remote bipolar channels. Each map averaged changes in two separate one-second windows and removed the channel mean. Direct transfer had a mean squared map error of 9.923 dB². A zero-change forecast had an error of 4.977 dB². The error ratio was 1.994, with a descriptive 95% patient bootstrap interval of 1.871 to 2.139. Transfer increased error in every patient. The longer-window check helped one patient but kept higher average error. Baseline drift maps also gave a ratio near two. Noise and changes between records can explain this result. The test does not measure a causal brain response or live BCI performance. The package preserves the fixed plan, all inputs, source hashes, code, complete results, and two figures. Fresh public files rebuilt all 280 stored arrays exactly. An independent Fourier calculation and a full numeric repeat passed.

stereoelectroencephalography · beta power · brain-computer interface · signal transfer · electrical stimulation