Two energy scores had 99 percent average agreement on top 10 percent SEEG lists with ten pulses per site in five patients

We tested whether subtraction of before-pulse energy removes the finite-pulse variance term in SEEG response scores. We used public OpenNeuro ds004457 version 1.0.2. The sample contained five patients, ten pulses per site, and 1,080 site and receiving contact pairs. We compared squared mean waveform energy with the mean product from distinct pulses. Both scores subtracted before-pulse energy and used the same observed before-pulse variance for normalization. The exact score difference was (V_after - V_before)/(10 V_before). Its mean patient median was 0.0120. The exact two-sided patient sign test gave p = 0.125. The top 10 percent lists had 99 percent average agreement across patients. Contrast signs changed in 66 pairs. The unnormalized product score is unbiased under a common mean and zero covariance between pulse errors. Normalization can cause ratio bias. Correlated errors and stable artifacts can remain. This five-patient sample does not establish a population effect or true neural connections. The package preserves code, input windows, source metadata, retrieval locks, proofs, figures, simulation draws, and all results. All 3,477,760 saved waveform bins matched the selected raw recordings exactly. Twelve numerical result files were reproduced byte for byte.

stereoelectroencephalography · electrical stimulation · pulse variance · waveform energy · distinct-pulse products · baseline subtraction · signal noise · measurement bias