Distinct pulse products gave a negative energy estimate for 44 micro-SEEG contacts in one pig at 20 to 200 ms

We tested four spatial modes in public OpenNeuro ds004819 version 1.0.0. One pig supplied eight current blocks, 80 fit pulses, and 79 later test pulses. The main set had 44 contacts with source impedance numbers at most 1000. This numeric rule is not a validated contact quality limit. We compared bases fitted to individual early pulses and to ten-pulse means. Products of different test pulses estimated repeatable energy at 20 to 200 ms after spinal stimulation. The pooled total estimate was -18,699 squared EDF-number units. Its descriptive 95 percent range was -65,046 to 27,252, conditional on fixed bases and blocks. Six blocks had negative estimates. Thus, the main test did not support a retained-energy fraction. All 64 contacts gave a positive total and four-mode ratios near one. Twenty contacts above the source-number threshold contributed 92.4 percent of that signed total under the 64-contact reference. This sensitivity does not identify valid neural contacts. A negative estimate does not prove absence of an evoked response. We did not identify a better fit rule. The package preserves code, exact inputs, source pins, signed results, three figures, and instructions. All eight raw rebuilds matched exactly. All values in eleven analysis outputs matched in a fresh process.

micro-SEEG · stimulation · repeat products · spatial compression · response energy · jackknife · pig · signal processing