A learned gate accepted most SEEG responses from excluded shafts in five offline sessions with two template pulses
A learned gate gave no clear detection gain in this offline test. We used public OpenNeuro ds004457 version 1.0.2. Five patients supplied 183 stimulation sites and 38 shaft folds. Each known site supplied two template pulses, one gate pulse, and one cutoff pulse. Four later pulses per site formed the test. The outer excluded shaft supplied no fitting voltage. Mean AUROC was 0.630 for the learned gate and 0.617 for fixed relative Mahalanobis distance. The mean patient gain was 0.013. Its descriptive 95% bootstrap interval was -0.019 to 0.044. The patient sign-flip test gave p = 0.50. The learned gate rejected 19.8% of unknown responses and 4.7% of known attempts. It accepted 80.2% of unknown responses. Calibration ended after 636 of 732 unknown test pulses. The result therefore has an offline scope. The package preserves data, code, 114 models, all scores, four figures, provenance, and reproduction instructions.
stereoelectroencephalography · brain-computer interface · novelty detection · Mahalanobis distance · evoked response
References
4- usesA fixed covariance rule gave higher SEEG site accuracy than a diagonal rule in offline tests from five patient sessions sci:sha256:1407cd9ac712a4bcd551a3738cc90c3fab309b871572cbd6fc644b41555f7886
We used its fixed covariance shrinkage rule for all response and background models.
- usesA relative distance score gave higher mean AUROC but accepted most SEEG responses from withheld shafts in five offline sessions sci:sha256:e20e4bb067514e9b579810b8a02215225203afff58108a2c2ada8ec7689b98d1
We used the preserved SEEG data extract, target-contact selection, source manifest, and excluded-shaft detection task. We tested a new learned gate with a matched two-pulse template budget.
- usesdoi:10.18112/openneuro.ds004457.v1.0.2
The analyzed five-session SEEG source and metadata were recorded as OpenNeuro ds004457 version 1.0.2.
- doi:10.1523/JNEUROSCI.1325-22.2023
Cited by
2 publicationsPublic catalog entries on Sciscape.
- 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 absent-condition rejection test and the offline timing audit as test concepts. The present score comparison uses a different dataset and condition definition.
- usesSixteen site templates gave lower mean AUROC for uncalibrated sites than four at 20 to 100 ms in five offline SEEG sessions
Its excluded-shaft rejection result informed the new capacity question. The present test varies template count with fixed known and uncalibrated test pulses. It does not replicate its learned gate.