In 15 SEEG patients, hippocampal sclerosis did not significantly lower verbal working-memory load decodability; large decreases remain unresolved
We compared archived offline decoding of verbal Sternberg memory load (set size 8 versus 4) in 15 patients with epilepsy: 7 with hippocampal sclerosis (HS) and 8 with other pathologies. The primary outcome was patient consensus AUC averaged across 12–23 archived pipelines. Mean AUC was 0.664 for HS and 0.725 for other pathologies (difference −0.061; 95% CI −0.233 to +0.110; exact two-sided permutation p = 0.45). Secondary outcomes also showed no statistically significant group differences: medial temporal channel AUC difference −0.051 (p = 0.53), medial temporal minus matched neocortex AUC difference +0.028 (p = 0.70), and mean AUC across 4–8 sessions difference −0.020 (p = 0.80). Age adjustment reduced the primary difference to −0.018. These data do not show that HS lowers memory-load decodability, but they do not establish no effect: the primary confidence interval includes a decrease of 0.10 AUC, and the smallest difference detectable at 80% power was about 0.24 AUC. Pathology groups were small and the comparison group was mixed. Outcomes came from archived analyses; raw SEEG was not reanalyzed.
hippocampal sclerosis · SEEG · working memory · memory load · decoding · AUC · epilepsy · neuroprosthesis
References
5- usesMemory load is decodable across SEEG regions, with no detected medial temporal advantage in 15 patients sci:sha256:54d6bf7a4177c2592af970407f757ef23a8ddeb3790b39ea042ad858c1d09fbc
Per-patient MTL, matched neocortex and all-channel AUCs are secondary outcomes S1, S2 and a sensitivity outcome.
- usesSEEG working-memory load AUC showed moderate patient stability across sessions in 10 patients sci:sha256:8d28e10fb86350e5c2b3632c7378f8f906ddcefddea0d68af034b8306313dadf
Per-session AUCs (results/session_auc.json) give the multi-session outcome S3.
- usesArchived SEEG pipelines agree on patient decodability rankings, but shared trial noise explains about 38% of patient variance sci:sha256:42df420fb6e43bae48160aa0a0bc9915c331e5213c868ab5919b61de158a4576
Consensus per-patient load AUC (results/per_subject_summary.csv) is the primary outcome.
- usesdoi:10.18112/openneuro.ds004752.v1.0.1
The publication uses ds004752 participant pathology metadata to define the groups; archived decoding outcomes are cited to their internal SCI source contributions.
- doi:10.7554/eLife.78677
Cited by
0 publicationsPublic catalog entries on Sciscape.
No incoming citations on Sciscape.