A sign test did not detect a late beta-to-alpha power ratio change at SEEG pairs at least 30 mm from stimulation in 15 patients

We tested whether the beta-to-alpha power ratio changed in late annotated windows around stimulation in OpenNeuro ds006519, version 1.0.0. We used the last three seconds of each annotation. Both endpoints of each included SEEG pair were at least 30 mm from both stimulation contacts. The primary set contained 907 channel instances in 22 recordings from 15 patients. Each patient contributed one value after aggregation by shaft and recording. The median ratio change was +0.280 dB. Its 95% bootstrap interval was -0.363 to +1.232 dB. Ten patients had positive changes and five had negative changes. The two-sided exact sign test gave p = 0.3018. The test did not detect a shared direction of change at the planned 0.05 level. This result does not establish equal ratios or a causal effect. Exact stimulation offset times and sham trials were absent. We assumed that exported voltages matched their bipolar contact labels. Thirteen recordings lacked stimulation coordinates and were excluded. All 35 source epoch rebuilds matched the saved arrays. A second raw reader also reproduced all 70 epochs. The package preserves code, source metadata, epochs, spectra, audits, patient results, figures, and a raw-data retrieval recipe with fixed hashes.

stereo-EEG · intracranial EEG · beta power · alpha power · power ratio · cortical stimulation · sign test