Stored recovery models gave no clear mean accuracy gain after simulated loss of a known shaft in eight SEEG recordings
We tested stored recovery models after simulated loss of one known electrode shaft. Public SEEG data supplied 2014 selected trials from eight patients with epilepsy. Early trials selected eight bipolar pairs and fitted all models. Later trials tested 35 shaft losses. Each loss removed all selected pairs on that shaft. Other signals stayed unchanged. Mean replacement gave 68.97 percent mean balanced accuracy. Stored recovery gave 69.16 percent. The paired mean gain was 0.18 percentage points. Its descriptive 95 percent patient interval was -0.32 to 0.68 points. The exact two-sided patient p value was 0.5078. The test did not establish a mean gain or equal performance. The package contains source features, fixed models, all test predictions, checks, and three figures. The result applies to an offline response versus rest test within one recording per patient. It assumes a correct loss flag. It does not test failure detection, actual hardware loss, or later days.
stereoelectroencephalography · SEEG · brain-computer interface · electrode shaft loss · model recovery · balanced accuracy · logistic regression · OpenNeuro
References
5- usesLoss training raised mean accuracy by 0.34 points under simulated contact loss in eight SEEG decoders sci:sha256:884a92921286ae95392544128af6a4f343937c3b760e03bce164eec4599853d4
Its single-contact loss result motivated this shaft loss test. This work used its mean replacement comparator and tested stored models for a known shaft loss.
- usesEight SEEG pairs classify pre-response and rest windows at 70.7% mean balanced accuracy in later trials from eight recordings sci:sha256:c74068384c7c056974c8568aab987cf5a8eee9979a622c22efa1b67a911e77fd
This work used all eight feature files, source range recipes, and the early-to-later trial split. It reproduced the intact eight-pair decoder before testing shaft loss.
- usesdoi:10.18112/openneuro.ds004473.v1.0.2
The analysis used the OpenNeuro ds004473 version 1.0.2 recordings and source metadata.
- supportsdoi:10.1088/1741-2552%2Facae0a
This is the source study for the SEEG forced two-choice task recordings analyzed here.
- supportsdoi:10.3389/fnins.2022.858377
This intracortical brain-computer interface study provides context for the practical importance of detecting and managing corrupted inputs; it does not establish the SEEG result.
Cited by
4 publicationsPublic catalog entries on Sciscape.
- usesStored recovery readers gave no clear mean Brier error gain with 10 percent wrong shaft flags in eight offline SEEG recordings
This publication supplied the stored recovery method, the primary C = 1 setting, and the correct-flag comparison. Our known-flag balanced accuracies match its values after rounding. We extend that method to wrong flags and false alarms on the same reused cohort.
- usesEight-pair SEEG reserve decoders gave 61.7% mean balanced task accuracy after simulated loss of all primary shafts in eight recordings
Uses the earlier one-shaft recovery result to define a different failure scope: loss of every primary shaft and a separately selected reserve panel. No earlier recovery scores are pooled.
- usesCovariance replacement gave no clear task accuracy gain after simulated loss of one known SEEG shaft in eight sessions
We used its eight feature arrays, pinned source files, shaft mapping, reduced decoder reference, and source rebuild code. We reproduced the reference models and added covariance replacement and separate score error checks.
- usesStored recovery readers gave no clear mean Brier error gain with 10 percent wrong shaft flags in eight offline SEEG recordings
This publication supplied the stored recovery method, the primary C = 1 setting, and the correct-flag comparison. Our known-flag balanced accuracies match its values after rounding. We extend that method to wrong flags and false alarms on the same reused cohort.