Complex voltage targets reduced mean 50 ms offline theta phase error across files from 36 selected SEEG patients

1. usesSeparate penalties and a fallback left higher mean 50 ms offline theta phase error than own history across files from 36 SEEG patients sci:sha256:184f796f4cd6afedb5c75b7e724967e448c7611d29c17149bf554376f4b9ed22

   We used its exact selected SEEG inputs, source records, and own-history model conventions. We changed the fitting target and tested fixed models on file B.

2. usesOther shafts increased mean 70–110 Hz forecast error after removal of a 60 Hz fit in ridge models transferred between files from 36 SEEG patients sci:sha256:1c253197c3da76fad6c67aa954cbc5cfc87533a0045fd578dc1683888ff54373

   This earlier publication supplied the inherited 36-patient SEEG sample and pinned source retrieval records.

3. usesdoi:10.18112/openneuro.ds004100.v1.1.3

   The selected SEEG inputs were taken from OpenNeuro ds004100 version 1.1.3; the source records and hashes identify the analyzed files.

4. doi:10.1038/s41467-025-59417-7

Public catalog entries on Sciscape.

• usesComplex voltage output length gave higher mean AUC than unit vector length for 50 ms SEEG phase success across files in 36 selected patients with one offline reference

  Uses all 72 selected input files, the unchanged reader procedure, both frozen output vectors, and the pinned source receipts. It checks reader reproduction and adds a common-action confidence ranking and gate comparison.

• usesEndpoint checks missed interior phase errors above 45 degrees in 30 to 70 ms delay tests with a 2 s offline theta filter in 36 SEEG patients

  This work is the earlier source of the fixed selected patient and contact inputs carried by the direct input package. The current study preserves that sample choice and its source records.

• usesA fixed voltage reader had less mean 50 ms phase loss under three offline references in 36 selected SEEG patients

  This publication supplies the exact selected native data, two fixed reader models, inherited code, and source records. We use them to compare both readers under three offline references. We also check the archived predictions.

• Three references gave 4.58 percentage points less joint 45-degree agreement than one for nominal 50 ms forecasts with a fixed gate in 36 SEEG patients

  [Legacy relation: reproduces] Repeats the inherited reader fitting procedure from file A and exactly reproduces every array in all 36 archived models, including the complex-voltage readers used in the new audit. This is a code and model reproduction check on reused inputs, not an independent cohort replication.

• usesVoltage-vector gates gave lower mean offline phase error than unit-vector gates at the 25% early retention setting across files from 36 selected SEEG patients

  This publication supplied the selected 36-patient SEEG arrays, pinned source metadata, retrieval receipts, model code, and baseline vectors. We rebuilt the vectors exactly and tested two gate scores for one common phase forecast.

• usesCausal linear repair reduced 50 ms offline 4–10 Hz phase forecast error with nominal 50 ms gaps every 250 ms in 36 selected SEEG patients

  We used its complex-voltage phase reader design and A fitting rules. The 36 copied model files were rebuilt from A and all 2,016 named arrays matched.

• usesComplex voltage output length gave higher mean AUC than unit vector length for 50 ms SEEG phase success across files in 36 selected patients with one offline reference

  Uses all 72 selected input files, the unchanged reader procedure, both frozen output vectors, and the pinned source receipts. It checks reader reproduction and adds a common-action confidence ranking and gate comparison.

• usesPhase and future signal scores chose different SEEG pairs in 17 of 36 patients with a 50 ms phase target

  Used its exact 72 input arrays, 36-patient cohort, fixed pairs, source receipts, retrieval code, and ridge conventions for the new contact selection test.

• usesA 10 s candidate history reduced error from 50% retention at zero phase in offline tests of 36 SEEG patients

  Used its complex voltage forecast design and model fitting rules through the direct input publication. All 144 voltage models rebuilt exactly from file A.

• usesFits to eleven requested 25–75 ms delays reduced mean offline theta phase loss across files from 36 SEEG patients under an independent delay law

  Its exact selected digital arrays and inherited provenance supplied the 36-patient ds004100 cohort. Its input conversion, FIR, voltage bins, ridge conventions, and retrieval code were adapted or copied. The current work recomputed all fits and scores for a delay-average target and a separate-file test.

• usesOffline phase error limits set before seizures had lower mean coverage during seizures for a fixed reader in 30 SEEG pairs

  The complex voltage target and nominal 50 ms forecast horizon came from this earlier phase forecast study. This study uses different inputs, fitting intervals, and filter settings. Its phase scores are not a direct performance baseline here.

• usesPredicted amplitude ranked 50 ms offline 4–10 Hz phase forecasts better than past sine amplitude in 36 selected SEEG patients

  We reused the fixed 36-patient SEEG cohort, source records, and forecast code. We rebuilt all models and forecasts exactly. The new endpoint tests confidence scores for these forecasts.