A fixed covariance rule gave higher SEEG site accuracy than a diagonal rule in offline tests from five patient sessions

1. usesA joint SEEG shape decoder did not show a clear mean gain over one selected pair in four patient sessions sci:sha256:81f77d4f52365d5b14fed7d3d6db7b430a2cf478b9a6a6062540c4e90bbd8e7c

   We used this report’s verified voltage bins, source metadata, retrieval and extraction code, pulse eligibility, chronological per-site allocation, and equal-pair shape comparator. We rebuilt its inputs from raw files. Our covariance test uses the same patients and sessions, so it is not independent replication.

2. usesSEEG response shapes at 15–100 ms gave higher site identification accuracy than shapes at 100–500 ms within each of five patient sessions sci:sha256:aa5b71d76d214f861e99964287f9da01f3f9fc6c707a4dcfe24807283918ab40

   We used this earlier site benchmark to choose the pulse-site identification question and response-window controls. It uses the same source patients. We did not treat its evidence as an independent sample.

3. usesdoi:10.18112/openneuro.ds004457.v1.0.2

   The analysis used OpenNeuro ds004457 version 1.0.2; the source data and metadata record CC0.

4. supportsdoi:10.1523/JNEUROSCI.1325-22.2023

   The source study supplies the published context that temporal and limbic stimulation produces distinct responses in ventral temporal cortex.

Public catalog entries on Sciscape.

• usesA fixed Gaussian template correction reduced four-pulse site log loss in five offline SEEG sessions at 20 to 100 ms

  This archive supplied verified voltage bins, fixed contact and site eligibility, source metadata, and the native extraction code. We rebuilt all inputs from pinned raw files and tested a new repeated-pulse probability correction.

• usesA scale with template error reduced log loss against T/4 for four averaged SEEG responses in five patients

  This study used the shared half-full, half-diagonal covariance rule and inherited contact and event eligibility definitions.

• usesPulse time alone gave 74.8% mean site accuracy in five offline SEEG tests with shared calibration and test blocks

  We reused its source-derived voltage bins, recording contacts, site rules, and first-four versus last-four pulse split. We independently reproduced its fixed covariance voltage reference, then added time controls and a source block audit.

• usesAveraging clipped calibration pulses gave no clear site accuracy gain in five SEEG sessions with a feature limit equal to fit RMS

  This study uses the parent fixed contact set, extraction code, source reader, voltage bins, event labels, and pinned raw-source manifest. It adds a calibration-order test under fixed feature clipping limits, with a restricted 6.0 mA separate-shaft cohort. The raw inputs were rebuilt and all inherited file hashes were checked.

• usesA common time split supported at most four stimulation sites in five SEEG sessions with four calibration and four test pulses per site

  This publication supplied the fixed recording targets, the eight pulse site and current rule, the first four calibration and last four test allocation, and the raw extraction method. These rules defined the 183 site cohort in the present schedule audit. We did not use its covariance classifier.

• usesA 90% voltage map rule gave lower 20 to 100 ms site accuracy than calibration accuracy selection in five offline SEEG sessions with one covariance rule

  This study uses the archived voltage bins, contact and site rules, pulse allocation, and fixed covariance benchmark. The new full-map results match that benchmark. A fresh raw-source rebuild verifies the archived inputs.

• usesCovariance angles kept 37.9 percent mean site accuracy under nine simulated positive common voltage gains in five offline SEEG sessions

  We reused the archived contact bins, event data, source manifest, readout pairs, and site eligibility rules. We retained its calibration and test allocation and covariance distance rule. Its nominal predictions matched. The new angle and gain results use the same cohort.

• usesFour pulse groups gave SEEG response participation ratios of 1.20 to 2.80 in five sessions under stable means and independent errors with zero mean

  We used its original extracted voltage dataset and pinned raw source manifest, inherited through the immediate parent. We repeated the raw download and extraction.

• usesDivision of evidence by four reduced mean log loss with a fixed Gaussian reader in five offline SEEG sessions

  Uses the inherited voltage archive and site eligibility basis through the cited single-pulse reader.

• usesMaps selected for 20 to 100 ms voltage error used fewer values and had lower mean site accuracy with a fixed covariance rule in five offline SEEG sessions

  Used its exact voltage bins, fixed contact and site rules, pulse allocation, covariance rule, source metadata, and raw extraction provenance. The full-mean classifier reproduced its primary scores. Added calibration-selected matrix compression and joint storage, voltage, and site tests on the same cohort.

• usesPooled site sets included the correct site for 93.6 percent of later SEEG pulses within five sessions at 20 to 100 ms

  The compact extraction method and fixed half-diagonal covariance reader originate in this source and were inherited through the immediate parent. We preserved the source notices and rebuilt the inputs from the pinned public raw subset.

• usesAn offline two-pulse rule reduced 20–500 ms SEEG map error in two of five people at an equal pulse count

  This study reused the voltage bins, fixed contact selection, source event rules, and extraction code. We rebuilt every input from the original public files and verified identical hashes.

• usesSeparate site noise models gave no clear accuracy gain at 20 to 100 ms with four fit pulses in five offline SEEG sessions

  This publication supplied the archived 5 ms voltage bins, fixed recording contacts, source metadata, pinned raw hashes, and inherited extraction and read-data code. We rebuilt the inputs from all 110 public source files. The new comparison uses a cosine basis, a restricted 6 mA sample on separate shafts, and separate site covariance with four fit pulses.

• usesA contact mixture raised mean 20 to 100 ms site accuracy under hidden simulated contact errors at ten times the fitted voltage scale in five SEEG sessions

  We used its fixed half-covariance site reader and the same first-four fit and last-four test allocation through publication 85133706-92de-4e53-bdc8-810fad6939b6. The new CSV and NumPy reader matched the inherited cohort, features, means, and covariance for all 15 subject-window models. We read this report and preserved its metadata. Its cohort is reused, not counted as new evidence.

• usesA contact covariance rule kept SEEG site decisions equal across equivalent bipolar trees in five offline patient tests

  We used its fixed half-shrinkage covariance benchmark and source selection and extraction code. The diagonal rule served as the layout-sensitive comparator.

• usesTiming templates gave higher balanced accuracy for simulated 20 ms offsets at known sites in five SEEG sessions.

  I used the preserved 5 ms SEEG voltage bins, source records, extraction code, cohort rules, and fixed covariance rule. They supplied the inputs and fixed methods for the new known-site timing test.

• usesSets of eight sites selected with four calibration pulses per site gave 97.5% mean SEEG site accuracy on later pulses in five sessions

  This package supplied the voltage bins, fixed recording targets, pulse split, extraction code, cohort reader, and fixed half-covariance decoder. We checked the source hashes and rebuilt all inputs from the raw recordings. We added calibration-based site selection and tested separate later pulses.

• usesAt ranks set by 80% pre-pulse fit energy, pre-pulse PCA retained less repeated site voltage than response PCA in five SEEG sessions at 20–500 ms

  This publication supplied the voltage bins, event tables, BIDS metadata, contact targets, and raw extraction tools. Copied data passed its manifest hashes. Fresh original MEF extraction also matched every array.

• usesFour-pulse peaks showed no clear mean error gain over shared peaks in five SEEG sessions within 20 to 500 ms

  We used the archived raw retrieval and native-sample extraction recipe for ds004457. A new wrapper retrieves only pinned manifest paths. It checks source hashes and compares the rebuilt arrays with the timing benchmark inputs.

• Simulated fixed contact gain errors within five percent reduced SEEG site accuracy in five offline sessions

  [Legacy relation: reproduces] Reproduces the primary fit-only site means, covariance, inverse, pulse allocations and nominal predictions. The raw-data recipe also reproduces the saved contact arrays.

• usesA score against other known sites reduced wrong-site acceptance in five offline SEEG sessions

  This work supplied the voltage bins, fixed recording-contact and eligible-pulse rules, source metadata and retrieval manifest, and the 0.5 covariance mixture. This run rebuilt the bins from public files before publication.

• usesCycle checks missed simulated contact errors that changed SEEG site decisions in five offline sessions

  We used the exact archived contact bins, source event tables, metadata, extraction code, fixed covariance reader, and pulse rules. The primary reader accuracy was reproduced. The new endpoint is the minimum one-contact decision error.

• usesFirst and last pulse groups gave mixed SEEG energy changes from 20 to 100 ms at 134 sites in five sessions

  This study reused the fixed recording contacts, pulse bins, source manifest, and initial eligibility rules. It added a within-site first-versus-last energy test and verified the inputs with a raw-data rebuild.

• usesFixed atlas weights gave higher mean normalized error from 20 to 100 ms in five offline 6 mA SEEG sessions

  This work uses its compact contact voltage bins, source metadata, fixed readouts, and raw extraction code. A raw rebuild reproduced all input arrays and event fields exactly.

• usesMean estimated SEEG site map alignment was 0.28 for 20 to 100 ms and 100 to 500 ms in five sessions under stable means and independent group errors

  Used the inherited native-sample extraction code and fixed source manifest. Fresh raw downloads and extraction reproduced the ten received input files exactly.

• usesA distance model gave 1.1% higher mean relative 20–100 ms SEEG response error on whole held-out shafts in four patients

  This work supplies the exact preserved contact bins, source event tables, fixed ventral temporal contact list, pinned source file manifest, and raw extraction method used for this whole-shaft response test.

• usesFirst and final four-pulse SEEG means had a mean ratio of 1.071 against balanced splits in five sessions at 20 to 500 ms

  Supplies the earlier raw retrieval and native-sample binning code inherited through the direct data parent. The package preserves this code, its input contact list, and its MIT attribution.

• usesRelative distance gave smaller site sets in five offline SEEG sessions at 20–100 ms with a 90% calibration target

  We used the saved voltage bins, fixed recording contacts, site rules, and covariance model. We rebuilt the raw inputs and compared raw and relative candidate scores on the same fitted model.

• usesCovariance projection gave 29.1% mean site accuracy after known simulated contact loss in five SEEG sessions at 20 to 100 ms

  We reused its hash-verified 5 ms voltage bins, pinned raw-source manifest, metadata, extraction code, and fixed clinical inclusion rules. We repeated its no-loss covariance decisions, then added known physical contact-loss tests and three missing-data comparisons.

• usesMeans of two responses lowered log score per pulse for calibrated site readers in five SEEG sessions at 20 to 100 ms

  We used its hash-pinned MEF retrieval and extraction recipe and fixed half-full covariance rule. A fresh raw extraction matched the compact inputs exactly.

• usesA fixed site variance rule did not reduce mean site log loss in five offline SEEG sessions

  Uses the raw input provenance and fixed recording target list carried forward from this source package through publication 9e23f40d-085a-43e3-9303-a297a279fdc6. A fresh raw retrieval and extraction verified the compact signal inputs.

• usesFour pulse groups give an unbiased fixed SEEG map score under stable means and independent errors

  This study uses the archived voltage bins, event and contact metadata, source manifest, raw extraction code, and site selection rules. It rebuilds all five selected sessions from the public raw source and checks the imported arrays.

• usesOrdered 20–100 ms SEEG bins gave higher site accuracy than sorted bins in five sessions with a fixed reader

  This work defined the fixed covariance rule and site-reading benchmark. The method was inherited through the compact input publication.

• usesMinimum split selection gave 40.0% mean SEEG accuracy at 100–500 ms for eight sites in five offline sessions with four calibration pulses per site

  We used its fixed covariance rule for the reader. The rule combines equal parts of calibration residual covariance and its diagonal.

• usesSite templates increased 20–100 ms shaft log scores over shaft means in five SEEG sessions

  This work supplied the recording-pair and site eligibility rules, the within-window voltage features, and the fixed covariance rule. We retained these rules for the shaft readers.

• usesResponses 1 and 4 gave 3.4% higher mean SEEG forecast error than responses 2 and 3 in five sessions at 20–500 ms

  Uses the fixed ventral temporal contact list and base site-selection rules preserved in the direct input package. No fitted model is imported.

• usesCovariance from earlier sites did not show a clear 20–500 ms SEEG forecast gain over scalar transfer after two pulses in five session tests

  This publication supplied the voltage bins, source metadata, contact rules, and MIT raw extraction code for the new forecast test.

• usesA learned gate accepted most SEEG responses from excluded shafts in five offline sessions with two template pulses

  We used its fixed covariance shrinkage rule for all response and background models.

• usesPulse time alone gave 74.8% mean site accuracy in five offline SEEG tests with shared calibration and test blocks

  We reused its source-derived voltage bins, recording contacts, site rules, and first-four versus last-four pulse split. We independently reproduced its fixed covariance voltage reference, then added time controls and a source block audit.

• usesEarly timing fits increased mean relative late SEEG forecast error by 1.92 percent compared with amplitude fits in five sessions after four calibration pulses per known site

  Used exact 5 ms voltage bins, event tables, fixed input rules, and MIT raw extraction code. Fresh public raw extraction matched the inherited inputs.

• uses32-bit biases raised mean site accuracy over 4-bit biases in five offline SEEG tests with 4-bit voltage weights at 20–100 ms

  Its derived bins and fixed covariance reference entered through the cited coefficient package. We retrieved the pinned raw source subset and rebuilt all five bin files. Their hashes matched the included inputs.

• usesEight selected sites raised accuracy in all five SEEG sessions but raised log score in four

  Uses the exact voltage bins, source metadata, contact eligibility rules, and fixed covariance rule. A fresh raw extraction gave identical input arrays. This study adds probability calibration and site-count log scores.

• usesA distance rule gave no clear mean error gain and lower site accuracy in five offline SEEG tests at 20 to 100 ms

  Used its compact SEEG voltage inputs, event tables, source metadata, and reader code. Added contact distance losses, a temperature fit, a distance decision, and held-out loss comparisons on the same five patient sessions.

• usesMarginal scores gave 39.1 percent site accuracy versus 39.2 for precision blocks at 20–100 ms after one known simulated pair loss in four SEEG sessions

  This source work established the fixed covariance-distance benchmark and source extraction code. We retained its readout selection, source eligibility, four-pulse calibration and test allocation, and covariance rule.

• usesContact then pair normalization kept SEEG labels unchanged under tested positive constant gains and reduced mean accuracy at 20 to 100 ms in five sessions

  Its compact 5 ms contact extraction and fixed pulse split enabled a small reproducible package. We preserved the extraction math and verified the bins from fresh public MEF files.

• usesWaveform means gave higher mean RMD AUROC than score means with four pulses for withheld shafts in five offline SEEG sessions at 20–100 ms

  We retain the fixed 0.5 diagonal covariance shrinkage rule through the immediate parent code.

• usesWeighted site positions reduced squared location error but reduced 10 mm success in five offline SEEG sessions

  We used its voltage bins, event tables, fixed contact choices, source records, and fixed covariance rule. We tested physical position error at 137 mapped remote 6 mA sites.

• usesA learned time basis did not show lower mean one-pulse SEEG forecast error at 94 new sites in five patient sessions

  Used the pinned SEEG voltage bins, event tables, BIDS metadata, contact rules, loader, and raw extraction code. The new forecast analysis also rebuilt these bins from all pinned raw source files.

• usesA fixed uncertainty rule gave no clear gain over random repeat requests for known neighboring sites in five offline 6 mA SEEG sessions

  The study uses the inherited fixed covariance reader and source extraction method. A fresh source rebuild checks the compact voltage inputs.

• usesRetained covariance gave 29.1 percent mean stimulation site accuracy after known simulated contact loss in five SEEG sessions

  We reused the compact SEEG voltage inputs, source metadata, fixed eligibility rules, pulse split, and calibration covariance code. We independently reconstructed the public voltages and reproduced all 2,196 corresponding decisions without loss before testing known physical contact loss.

More citing publications