Selection on two shafts increased SEEG shape retention after simulated worst shaft loss at 82 sites in eight recordings
Selection on two electrode shafts increased later SEEG shape retention after simulated complete loss of one shaft. Both rules selected two bipolar pairs with four distinct contacts. Five early pulses set the selection and fixed templates. The test used 82 stimulation sites and 546 later pulses in eight public epilepsy recordings. The source was OpenNeuro ds004696 version 1.0.1. A shape match required signed correlation of at least 0.8 in the 20 to 500 ms window. Evaluation removed the least favorable fixed shaft at each site. Retention increased from 27.22% to 29.83%. The equal person mean gain was 2.61 percentage points. Three people had a gain. Five people had no change. With all contacts present, retention decreased from 69.94% to 69.05%. The direction of the loss-test gain is structurally protected by the selection rule under the stated assumptions. The measured gain does not establish a population benefit or live brain-computer interface performance. The package preserves exact inputs, source hashes, retrieval code, results, three figures, and reproduction instructions. Independent checks passed. A clean replay gave identical numerical files. A fresh source download reproduced one person exactly. The source checks for the other seven people are inherited.
stereoelectroencephalography · SEEG · shaft loss · shape retention · electrical stimulation · brain-computer interface
References
5- usesAt rate 1, learned weights reduced 20 to 500 ms SEEG forecast error against equal weights but raised error against scaled means in eight recordings sci:sha256:ef0c0959114fb72eaef63b2306de27883d737af4882e1642fc43cc265cd97c7b
This publication supplied the exact all-pulse Gram archives, native voltage example, source metadata, retrieval helpers, source hashes, and inherited audit records. This study tests a new four-contact shaft-loss shape endpoint on the same cohort.
- usesComplementary pair selection gave 68.4% later SEEG shape matches versus 68.8% for separate ranking in eight sessions at r ≥ 0.8 sci:sha256:31784984be345bd06923be14f99677f0cd684aa733f48a90c69148caf7df04f9
The four-contact pair comparison and signed shape-match endpoint motivated this study. This study tests complete shaft loss and all retained later pulses. It does not claim a new general pair-selection principle.
- usesSelection for retained template energy gave no clear 100 to 500 ms AUC gain after known simulated shaft loss in seven SEEG sessions at eight contacts sci:sha256:14b52e62c5b70ffcb50cd41f2950dcf5b3030d62a86de2943d6c16d7afae537b
This earlier work tested selection for retained template energy after shaft loss in the same source cohort at eight contacts. It supplies prior method context. The present endpoint uses signed shape matches at four contacts. This publication was reviewed after the primary analysis; no selection or endpoint changed.
- usesdoi:10.18112/openneuro.ds004696.v1.0.1
The analysis used source trial voltage derivatives from OpenNeuro ds004696 version 1.0.1.
- doi:10.1523/JNEUROSCI.2201-22.2023
Cited by
3 publicationsPublic catalog entries on Sciscape.
- usesNested pair selection did not show lower absolute score error at 83 SEEG sites in eight recordings
We used its exact all-pulse waveform inner-product matrices, native voltage example, event tables, and source provenance. These inputs support the chronological and random-partition validation tests.
- usesCovariance from remaining inputs increased mean SEEG accuracy for eight sites from 27.1% to 34.7% after known simulated shaft loss in three offline sessions at 100 to 500 ms
Its complete-shaft-loss benchmark informed the choice to remove every recording feature on one shaft. This study uses a different dataset and a site-identification endpoint. We used no records or numerical results from that study.
- usesCovariance from remaining inputs increased mean SEEG accuracy for eight sites from 27.1% to 34.7% after known simulated shaft loss in three offline sessions at 100 to 500 ms
Its complete-shaft-loss benchmark informed the choice to remove every recording feature on one shaft. This study uses a different dataset and a site-identification endpoint. We used no records or numerical results from that study.