Selection for distinct response patterns had an uncertain mean accuracy gain at four SEEG pairs in eight STReEF patients

1. usesThe nearest half of SEEG pairs had a higher mean rate of repeatable waveforms across recorded contact order reversal in eight STReEF patients sci:sha256:399c1198c4184a27bd9be26923336885a19d4fdd7d9a5d49a4ca8f9cffac5cae

   We used its fixed STReEF metadata sample, source metadata, object version pins, byte range receipts, and MIT retrieval code. We selected its first four conditions per patient. We recovered the voltage again and computed new features, selection rules, and test results.

2. supersedesSelection for distinct response patterns had an uncertain mean accuracy gain at four SEEG pairs in eight STReEF patients sci:sha256:5edddca3e74413468a090e57980eb8e55c89f799e4b752d4de9edfdf5ff890b5

   This version corrects a display rounding mismatch. It states the exact main score, 122/160 or 76.25 percent. The voltage features, selected pairs, predictions, and uncertainty estimates are unchanged.

3. usesdoi:10.18112/openneuro.ds005448.v1.0.0

   The analysis used STReEF OpenNeuro ds005448 version 1.0.0; the recorded source DOI is version-specific.

4. supportsdoi:10.1016/j.ynirp.2025.100274

   The STReEF article describes the clinical source recordings and dataset context used in this analysis.

Public catalog entries on Sciscape.

• usesTwo calibration pulses gave lower mean four-condition accuracy than five at four SEEG pairs in eight STReEF patients after recorded contact order reversal

  We reused its complete four-condition feature sample, test pulses, fixed reader and pair-selection rules, source receipts, and five-pulse result. The new analysis changes calibration input count and adds all early two-pulse subsets and component diagnostics. Independent checks reproduce its five-pulse counts exactly.

• usesSurviving contrasts raised mean 20–100 ms condition accuracy from 71.24% to 73.15% after known simulated contact loss in eight SEEG patients with unchanged surviving signals

  This work supplied the fixed STReEF sample, calibration selector, features, source records, and retrieval code. All 32 selected layouts and 640 nominal decisions matched. The new test uses known simulated physical contact loss.

• usesFive templates per site certified 66.46% of SEEG search labels in eight patients on a ±9.77 ms grid under exact arithmetic

  Used the metadata selection and pinned raw-range retrieval helpers through the immediate input publication. The fixed four-site sample and early/later split originate in this work. This is the same cohort, not an independent replication.

• usesSelection for one known contact loss gave lower mean minimum site accuracy in eight offline SEEG tests with four disjoint pairs at 20–100 ms

  This work supplied the fixed four-site sample, all eight-patient features, the feature extractor, source checks, and MIT code. We kept the inputs, rebuilt all 320 public raw ranges, and reproduced its distinct-selection reader as an input check. The new comparison uses exact selection under an equal eight-contact budget and known contact omissions.

• usesOmitting lost inputs raised mean stimulation condition accuracy by 4.8 points in eight offline SEEG tests with known simulated contact loss

  We reused its complete four-condition feature sample, five calibration and five test pulses per condition, fixed four-pair reader, source metadata, range receipts, and MIT retrieval code. We retrieved all 320 pinned epochs again and matched every array exactly. The fit, selected pairs, and intact decisions match its archive. We added known contact and shaft loss tests.

• usesAbsolute distance had no clear mean AUC gain over the label gap for absent conditions at four SEEG pairs in an offline test of eight patients

  Uses the fixed four-condition sample, voltage features, source metadata, native byte examples, and retrieval code. These inputs permit a matched score comparison with no voltage from the absent condition in fitting or calibration.

• usesFive-pulse products raised mean log loss in eight SEEG patients versus geometric means after recorded contact order reversal

  The four-pair selection rule comes from this earlier work. The present reader preserves its greedy log determinant rule and fits it with training pulses only.

• usesCalibrated Student t scores gave no observed mean loss gain in eight SEEG patients after contact order reversal with four recording pairs

  We used its four-pair greedy log determinant rule, diagonal variance rule, eight cosine coefficients, and retrospective candidate-pair definitions. Each calibration fold refitted the pair rule without its held-out pulse.

• usesDisjoint SEEG pairs gave 53.1% mean minimum four-site accuracy, compared with 53.8%, in eight STReEF patients. Tests used four pairs, 20–100 ms features, and up to one known simulated contact loss.

  The direct parent used this earlier fixed four-site sample and fit-only separation rule. We retained the sample and fit score for the contact-loss comparison.

• usesGlobal amplitude removal lowered mean site accuracy after recorded contact order reversal in eight STReEF patients with four SEEG pairs

  We used its four-site sample in each of eight STReEF SEEG patients. We reused its compact features, source records, and retrieval code. We rebuilt all 320 native epoch ranges. The new analysis compares amplitude and normalized site readers.

• usesA four-pair SEEG reader kept 73.1% four-site accuracy under 16 ideal recording sign patterns in eight patients

  We used its four-site sample in each of eight STReEF SEEG patients through the direct input publication. The condition order, fit and test pulse groups, and source feature construction remained fixed.

• usesA four-pair SEEG reader kept 73.1% four-site accuracy under 16 ideal recording sign patterns in eight patients

  We used its four-site sample in each of eight STReEF SEEG patients through the direct input publication. The condition order, fit and test pulse groups, and source feature construction remained fixed.

• usesA gap gate kept 24 of 160 SEEG response labels correct across a simulated ±9.77 ms grid in eight patients

  Used the inherited metadata-selected four-condition sample and four-pair model specification through the timing parent. No new gate result selected the cohort.

• usesFive late SEEG bins had lower mean accuracy than one window at equal value and pair counts in eight STReEF patients

  This study uses the distinct-pattern pair selector, pooled noise scale, and nearest-mean reader to compare two late feature sets at equal counts.

• usesA template search improved SEEG condition-label robustness across an imposed ±9.77 ms error grid in eight patients

  We reused its fixed four-condition sample, public source receipts, source metadata, and metadata check helpers. We tested a new response window and a search over shifted templates.

• usesA voltage bound for one contact certified five more correct labels in eight offline SEEG readers at a fit voltage benchmark with fixed references

  Uses the four-condition pulse sample and pair selection rule inherited through the immediate source. No new pairs or pulses are selected with test results.

• usesTwo calibration pulses gave lower mean four-condition accuracy than five at four SEEG pairs in eight STReEF patients after recorded contact order reversal

  We reused its complete four-condition feature sample, test pulses, fixed reader and pair-selection rules, source receipts, and five-pulse result. The new analysis changes calibration input count and adds all early two-pulse subsets and component diagnostics. Independent checks reproduce its five-pulse counts exactly.