Joint label changes preserved shared task labels and class counts in 944 SEEG trials from 12 patients
The number task in OpenNeuro ds005415 shares its labels across patients in each of two task versions. We built a joint label test for 944 trials from 12 SEEG patients. In 9,999 draws per scheme, it kept shared task labels and each patient's format and class counts. A fixed trial-order decoder gave mean AUC 0.546. Its upper-tail p was 0.0555 with independent patient labels and 0.3548 with joint labels. A five-block joint test gave p = 0.2098. These tests assume label exchange within their stated strata. The SEEG power score was 0.488 at published event times. Reused raw timing probes support a possible signal-origin error, so this score cannot establish a neural conclusion. The package preserves the method, features, predictions, source versions, hashes, timing evidence, and code. A source rebuild reproduced all 12 feature files. A final offline run matched all 74 expected output hashes.
stereotactic EEG · quantity decoding · joint label permutation · exchangeability · event alignment · numerical cognition
References
3- usesVersion 1.0.0 of ds005415 has an EDF length mismatch and an event sample rate mismatch sci:sha256:90c2c21bd2c5e0ed3301dd693d4743ee0a6103a4c8783bdd8bd6c2c22bd9f61b
We used 30 archived raw timing probes and their source records to check the reported event-to-signal origin concern. A separate parser reproduced all 120 auxiliary pair measurements. This evidence limited the neural interpretation of our decoder scores. We added a joint label method for the shared task design.
- usesdoi:10.18112/openneuro.ds005415.v1.0.0
The analysis used SEEG signals and task tables from the exact OpenNeuro ds005415 version 1.0.0 snapshot.
- doi:10.1371/journal.pone.0313155
Background source study for the Numbers task and dataset; this publication uses a different analysis and does not reproduce or refute its full analysis.
Cited by
2 publicationsPublic catalog entries on Sciscape.
- usesA fixed 70 to 110 Hz decoder did not detect number transfer from nonsymbolic to symbolic SEEG trials in four records under uniform sample timing
Adapted its shared task-slot label test to five number classes. Preserved the shared sequence and all patient-by-stimulus-type-by-half class counts.
- usesShifted event times gave 53.8% mean sensory mode transfer accuracy in four SEEG records under nominal sample timing
Its shared task label result informed our joint sensory mode null test. We preserved shared slot labels within each task version. We also conditioned on exact patient and half membership to preserve class counts.