Average references reduced low frequency seizure onset ranking AUC in two short SEEG recordings from each of 36 patients

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

   The analysis used OpenNeuro ds004100 version 1.1.3; the fixed Git metadata commit and source object versions and hashes are preserved in the package.

2. doi:10.1038/s41598-019-42347-y
3. doi:10.1093/braincomms%2Ffcab231

Public catalog entries on Sciscape.

• usesLower contact numbers ranked source onset marks with mean AUC 0.671 in 36 SEEG patients

  Uses all 72 archived window-power files, exact contact mask, source channel metadata, extraction code, and pinned retrieval records. The complete input publication is copied and verified. All 108 patient-by-reference signal rows are reproduced. The new tests compare contact number with the fixed signal and use patient-held-out rank models.

• usesA low frequency rank threshold gave 61.5 percent balanced accuracy in 36 SEEG patients when each test patient was excluded from fitting

  We reused 72 band-power arrays, contact labels, source metadata, version-locked retrieval records, and EDF extraction code. We also used its contact score table for an exact comparison. These inputs enabled the new patient threshold transfer test.

• usesEarly bipolar gamma change maps had more repeat agreement than baseline change maps after power and position controls in two seizure files per patient for 28 SEEG patients

  The preserved input extraction uses its versioned range transport and EDF layout methods. We used the inherited range transport for the fresh source check and a separate NumPy signal reader for power calculation.

• usesA fixed logistic model had lower mean seizure onset ranking AUC with two added power variation features in 18 SEEG test patients

  This publication supplied window powers, source metadata, channel rules, extraction code, and pinned EDF retrieval records. We preserved its source files and independently rebuilt 15 windows from each of two raw records.

• usesTop ten percent gamma lists had 42.7 percent mean agreement across two seizures with a rank rule over five starts and 46.2 percent with one window in 28 SEEG patients with common average reference

  Used the inherited range retrieval and EDF record layout method through the preserved source code.

• usesA shared baseline increased early to late gamma change agreement in 28 ds004100 SEEG patients with common average reference

  The source extraction and independent source audit use its HTTP range retrieval code. That code retrieves fixed public EDF byte ranges from pinned S3 object versions.

• usesBaseline division lowered early gamma AUC against onset labels in 32 SEEG patients

  The inherited extraction uses this public S3 range transport and EDF layout method. The current raw audit also executes that transport code.

• usesMinimum rank selection gave no detected precision change at a 10 percent count in 36 patients' two 120-second SEEG records under independent symmetric effects

  We used all 72 preserved window-power files, the matched contact set, metadata, extraction code, and pinned raw retrieval records.

• usesA fixed low frequency score had higher mean contact rank agreement within runs than across runs in 56-second samples from 36 SEEG patients

  This secondary analysis uses all 72 archived window-power feature arrays, source contact metadata, matching rules, source hashes, and pinned raw retrieval code. It adds equal-duration within-run and across-run rank comparisons in the same 36 patients.

• usesEarly gamma ranks had higher mean agreement across references than across seizures in 32 ds004100 SEEG patients

  We used its S3 range transport and EDF record layout method to retrieve public seizure windows. The package preserves the prior source code. The new code adds fixed onset windows and matched contact pairs.

• usesBaseline division lowered mean early gamma AUC for seizure onset labels in two seizures from each of 32 SEEG patients under a common average reference

  The new complete signal audit reuses the public S3 byte-range transport inherited through the selected power archive. That transport checks range sizes and pinned object versions. This study adds independent EDF decoding and a new comparison of early gamma scores. The original transport code and its publication metadata are preserved inside the source archive.

• usesThe 112-second SEEG test did not detect better onset ranking across two records in 36 patients under independence and null sign symmetry

  Used all 72 window-power files, the matched contact set, source metadata, and pinned retrieval records. These files enabled the equal duration sampling test. The complete source package is preserved unchanged.

• usesEarly bipolar gamma change had higher local order agreement than pre-onset change across two seizures in 31 ds004100 SEEG patients

  The preserved original extractor uses the EDF byte-range transport method from this publication. Its pinned retrieval records permit raw-source verification without full-recording downloads.

• usesGamma power change had higher mean clinical onset mark AUC at 16 to 20 seconds than at 0 to 4 seconds in 28 ds004100 SEEG patients with common average reference

  The inherited extractor uses this work's versioned EDF byte-range retrieval method. We preserved the code and fixed source range recipes.

• usesLower contact numbers gave mean seizure onset AUC 0.794 within mixed inferred shafts in 34 selected ds004100 SEEG patients

  Its reference-sensitive onset ranking and within-shaft power comparisons supplied prior context. We did not use its 120 s signal arrays.

• usesUnder patient independence, the onset AUC change interval for added baseline gamma included zero in a fixed common average SEEG model with 16 test patients

  The inherited extraction uses the fixed EDF byte-range method from this source. Our raw audit used its inherited transport code, with separate edfio decoding and SciPy spectral calculation.

• usesGamma change raised mean AUC for onset marks after high gamma control in fixed models for selected pairs from 16 SEEG test patients

  The preserved upstream extractor used the range retrieval method from this work. Its source implementation is included for provenance.

• usesInterictal gamma gave an uncertain gain in early seizure rank prediction after baseline adjustment in 29 ds004100 SEEG patients

  We used the range transport and EDF layout method preserved in the seizure parent. The new source checks use the copied transport helper to retrieve fixed signal bytes.

• usesBaseline division reduced mean early gamma AUC within named electrode shafts in 29 SEEG patients under a common average reference

  The source extraction and current raw audit use its EDF byte-range transport through the immediate source publication. The copied transport code and source metadata are preserved.

• usesA low frequency score had 0.058 excess mean AUC under uniform shaft label permutations in 72 short SEEG records from 36 patients

  The archived window-level band densities, clinical labels, channel names, and pinned source records enabled the new exact shaft null, paired label permutations, circular shifts, and adjacent contact comparisons. The full input package is preserved unchanged.

• usesMinimum rank selection gave no detected precision change at a 10 percent count in 36 patients' two 120-second SEEG records under independent symmetric effects

  We used all 72 preserved window-power files, the matched contact set, metadata, extraction code, and pinned raw retrieval records.

• usesA shift based on training labels reduced probability error and kept contact order in a fixed SEEG model tested on 36 patients

  This publication supplies the unchanged 72 SEEG window-power files, clinical labels, fixed 36-patient contact set, source metadata, and pinned raw retrieval records. These inputs support the new probability error audit.

• usesA separate seizure baseline lowered mean early gamma onset-label AUC in 32 SEEG patients under common average reference

  The inherited extraction and raw audit use this work's S3 range transport and EDF layout methods. Their preserved code permitted a fixed-version raw signal audit without full EDF downloads.

• usesA fixed low frequency score added 0.034 to mean onset-label AUC above a coordinate and number model in 36 SEEG patient holdouts

  The 72 stored-reference power files, source selection, immutable EDF object references, and extraction code supplied the signal inputs. This study adds original source coordinates, a fixed location and contact-number control, patient holdouts, and patient bootstrap model refits. The patients are the same; this is not an independent cohort.

• usesThe patient interval for gamma rank gain after onset included zero at 70 boundaries in 29 SEEG patients under independent, exchangeable sampling

  The inherited extractor and new raw verification use the EDF byte-range transport method from this publication. Pinned public ranges permit source checks without full-recording downloads.

• usesThe 95% interval for SOZ AUC gain from patient ranks included zero for a fixed logistic model in 18 test SEEG patients

  We rebuilt the model inputs from its 72 saved window feature files. Its metadata and versioned signal retrieval records enabled fresh checks of four public EDF excerpts.

• usesBipolar gamma power had mean clinical onset mark AUCs of 0.900 at 0 to 10 seconds and 0.883 at 10 to 20 seconds within named shafts in 26 ds004100 SEEG patients

  Its versioned HTTP byte-range method supports the preserved source retrieval recipes and optional fresh source checks.

• usesFree label shuffles gave 15.2 percent rejection under a circular label null in SEEG score maps from 36 patients

  We used its full window power archive, contact selection, source metadata, signal extractor, and range locks. We rebuilt all 72 pinned prefixes and checked every saved array. Its full package is preserved under upstream.

• usesMean early gamma ranks from two seizures gave 27.5 percent precision for onset marks under a ten percent contact count rule in 32 ds004100 SEEG patients

  Uses the inherited fixed-version range transport method for the fresh raw signal audit of all 64 recordings.

• usesSmaller scaled contact numbers ranked seizure onset labels with mean AUC 0.676 in two short SEEG records for each of 36 patients

  We used all 72 checked SEEG feature files, the source metadata, and the pinned source retrieval manifests. We added a scaled contact number baseline, models tested on held-out patients, restricted pair lists, and independent numerical checks. This study uses the same cohort and is not an independent clinical validation.

• usesA fixed low frequency score ranked seizure onset contacts less well at named contact boundaries in 34 SEEG patients

  The preserved window powers, source records, pinned retrieval recipes, and matched contact set supplied the data. New code tested immediate named opposite-label neighbors and a paired local-versus-all comparison.

• usesA 56-second SEEG threshold test on known contacts did not detect a mean balanced accuracy loss between recordings in 36 patients

  This source supplied the 36-patient cohort, window power values, labels, matched contact rules, measurement code, metadata, and pinned signal retrieval records. The complete source package is preserved and all 341 archived file hashes were checked.

• usesSeparate binary labels for mixed contact pairs gave a mean AUC bound width of 0.098 for a fixed score in 36 SEEG patients

  This study reuses the 36-patient selection, 72 source locks, source metadata, EDF reader, spectral routine, and fixed score direction. It retrieves the same raw prefixes and adds pairs with different contact onset labels. All 320175 powers for pairs with equal labels match the prior values exactly. The new study averages file scores before AUC, so it does not reproduce the prior mean of file AUC values.

• usesSelection of contacts in named adjacent pairs with different onset labels reduced mean AUC by 0.130 for a fixed low frequency score in an exploratory test of 34 SEEG patients

  This archive supplied all 72 feature files, clinical metadata, pinned raw object records, and extraction code. We used them to test adjacent contacts and repeat the public raw extraction.

• usesStored low frequency SEEG scores exceeded a fixed shaft count null by 0.058 mean AUC in two 120-second recordings per patient across 36 patients

  We used its matched contact table, window powers, fixed score, source metadata, and pinned raw retrieval recipe. These inputs enabled the exact shaft count AUC benchmark and paired conditional tests. All parent files were verified and preserved. The new result uses the same cohort and score.

• usesA low frequency SEEG score exceeded a control with fixed shaft onset counts by 0.058 mean AUC in two 120 s interictal records from each of 36 patients

  This publication supplied all 72 window-power feature files, fixed low frequency score definitions, the 36-patient matched cohort, clinical labels, and pinned raw signal retrieval records. We verified its full file manifest, reconstructed its scores, and used these measurements to calculate the new exact control and patient contrasts.

• usesA contact number rule ranked adjacent seizure onset contacts better than a fixed power score in two short SEEG recordings from each of 34 patients

  This study uses the exact 72 window-level feature archives, stable clinical labels, source metadata, and pinned retrieval records. It reproduces the source full contact AUC. It extends the known within-shaft check to adjacent opposite-label pairs and a fixed contact number rule.

• usesA low frequency SEEG score had higher mean rank agreement within records than between records in 36 patients with 56 seconds of signal per sample

  This work supplied all 72 spectral feature files, the fixed 36-patient cohort, matched contact rules, extraction code, source metadata, and pinned raw signal retrieval records. We used them to compare separate equal-size samples within and between records and to measure agreement at several signal amounts. We also reproduced the original full-record rank correlation point estimates.

• usesIn 28 SEEG patients, minimum rank across five window starts gave mean separate seizure rank AUC 0.750 versus 0.762 for onset selection at a ten percent quota

  The current signal check uses the inherited range transport and EDF layout code lineage preserved in the original signal study. Exact object versions and range hashes permit a repeat check of the used powers.

• usesTop ten percent gamma lists had mean retention of 51 percent across five window starts with a fixed baseline and common average reference in 28 ds004100 SEEG patients

  The new signal retrieval uses the range transport inherited through the direct source archive. The preserved code identifies this method source.

• usesA fixed spectral model predicted random patient labels above chance across separate contacts and named shafts in 36 known SEEG patients

  Used the archived 72 record spectral measurement files, fixed metadata, and pinned raw retrieval code. Fresh downloads matched all 72 prefix hashes. Feature arrays agreed within the stated numeric tolerance.

• usesFull baseline correction increased mean error for centered early gamma maps in 32 SEEG patients with two seizures each and common average reference

  We used its inherited byte-range transport code in the fresh raw audit. A separate EDF decoder and power calculation checked all stored source powers.

• usesMean ranks gave 25.2% patient mean onset label recall with 10% contact budgets in 36 patients with two SEEG records

  This source supplied all 72 window power files, labels, extraction code, metadata, and pinned source versions. These inputs enabled the fixed contact budget study.

• usesRestriction to boundary contacts reduced mean low frequency score AUC from 0.648 to 0.518 in two 120-second SEEG samples from each of 34 patients

  This source supplied all 72 SEEG feature files, extraction code, clinical channel labels, and pinned retrieval records. Its known global and within-shaft results helped select the original boundary question. The full source package is preserved unchanged.

• usesEarly gamma change had lower mean local agreement with onset marks than baseline gamma power in common average SEEG signals from 30 patients

  Uses its inherited S3 range transport and EDF record-layout method. The source code and its later onset-window adaptation are preserved. Its interictal results are not new study inputs.

• usesEarly gamma increased mean onset label AUC by 0.067 beyond baseline power and contact position in 16 ds004100 SEEG test patients

  The source extractor and current raw signal audit use its public S3 range transport and EDF record layout method through the cited source package. The inherited original code and publication record are preserved.

• usesA fixed spectral random forest had higher mean AUC with contact tests than patient tests in 36 SEEG patients

  The archived spectral measurements and fixed contact cohort supplied inputs for new contact, shaft, and patient test comparisons.

• usesAdding power variation lowered mean seizure onset ranking AUC in a fixed linear model of two short SEEG records from each of 36 patients

  We used its 72 window-power files, matched contact set, pinned dataset metadata, source code, and retrieval records. We verified all 341 archived source files and preserved them without edits.

• usesThe gamma timing score gave no clear onset mark AUC gain over power and position in 14 test SEEG patients

  We used its fixed source object and byte range method. This method allowed source checks without full EDF downloads.

• usesBaseline division lowered early gamma onset label AUC in 32 ds004100 SEEG patients with common average reference

  The reused extraction code uses its versioned EDF byte-range transport. We preserve this helper and use it for the fresh source check with separate signal decoding and spectral calculation.

More citing publications