sciscape Publications
← All publications

A fixed 65 to 95 Hz SEEG decoder gave a mean transfer AUC of 0.475 between Arabic digits and spoken numbers in eight patients

We tested number class transfer between Arabic digits and spoken numbers with public SEEG data from OpenNeuro ds005415, version 1.0.0. A fixed 65 to 95 Hz power feature and ridge decoder used 469 trials from eight patients. The mean transfer AUC was 0.475. The 95 percent subject bootstrap interval was 0.421 to 0.528. A blocked label test with 9,999 permutations gave p = 0.7487. A secondary stimulus format control also failed. The result applies to this decoder and sample. It does not show that the brain lacks a shared number code. The package preserves trial features, code, results, source versions, byte range hashes, metadata, and retrieval instructions. Source checks found a short EDF object, failed cue intervals, and an event sample time scale that differs from the SEEG rate.

stereoelectroencephalography · number cognition · cross-format transfer · ridge regression · area under the curve · negative result · OpenNeuro

Admitted 0 citing publications

References

2
  1. usesdoi:10.18112/openneuro.ds005415.v1.0.0

    The analysis used SEEG signals and metadata from the Numbers dataset, OpenNeuro ds005415 version 1.0.0.

  2. doi:10.1371/journal.pone.0313155

Cited by

0 publications

Public catalog entries on Sciscape.

No incoming citations on Sciscape.