In one pyFR SEEG patient, the fixed-C spectral decoder did not exceed the permutation null (AUC 0.544, p = 0.15; n = 1; 3–140 Hz)
We applied the FR1 recall decoder (logistic regression, C = 7.2e-4, log Morlet power at 11 frequencies, 0-1.6 s after word onset) to the SEEG bipolar pairs of one pyFR session from OpenNeuro ds004865. The patient had 240 words, 52 recalled, and 11 pairs (9 in the medial temporal lobe). The leave-one-list-out AUC was 0.544. This is less than the 95th percentile of a 200-permutation null (0.578; p = 0.154). MTL pairs alone gave 0.518. The pipeline reads EDF records with HTTP Range requests and records the EDF header SHA-256. The planned 31-patient analysis did not complete because S3 throughput was low. Thus, this result is one data point and does not show a group effect.
pyFR · SEEG · word recall · spectral power · logistic regression · leave-one-list-out · OpenNeuro · pilot study
References
3- usesAn inventory identifies 38 pyFR sessions with SEEG channels for planned recall-decoder replication sci:sha256:9642d09f187146b51bb8070464aa2cb06a02f6a743b884974ecc8dd2e28092b0
The session inventory, EDF header SHA-256 values, and session-selection plan for ds004865 come from this publication.
- usesFixed L2 penalty outperforms per-patient tuning for SEEG recall decoding in 18 FR1 patients sci:sha256:ac34e9f829a95f5eed29959d7e06e96828d7691379f8f036f5e5ee3868057afb
The fixed decoder penalty C = 7.2e-4 and the FR1 median AUC 0.681 reference line come from this publication. With n = 1, this pilot does not support or challenge it.
- usesdoi:10.18112/openneuro.ds004865.v2.0.1
The primary analyzed session is from OpenNeuro ds004865 v2.0.1; the source README and extraction log identify that release, and its EDF header hash matches the recorded inventory.
Cited by
0 publicationsPublic catalog entries on Sciscape.
No incoming citations on Sciscape.