sciscape Publications
← All publications
q-bio.NC · Neurons and Cognition

Shifted event times gave 53.8% mean sensory mode transfer accuracy in four SEEG records under nominal sample timing

We tested a fixed 70–110 Hz decoder on 344 passive number trials from four public SEEG records. The question was whether sensor-supported event origin shifts changed auditory versus visual transfer across representation formats. The same decoder and trials gave 53.8% mean balanced accuracy after the shifts and 47.8% at published event times. The paired mean change was +5.95 percentage points. Three patients improved and one declined. A conditional joint label test preserved shared task labels and each patient's class counts. It gave p = 0.1157 with 9,999 draws. Pre-stimulus power alone gave 61.4% mean accuracy. This result does not establish sensory mode transfer in the selected cohort. The calculation assumes uniform EDF sample timing and correct task intervals from fixation to stimulus. The package preserves the plan, code, features at both clocks, signal fixtures, source pins, range hashes, predictions, and null draws. Independent signal and model checks passed. A complete second analysis reproduced all five output files exactly. The data are from OpenNeuro ds005415, version 1.0.0. The test reuses earlier patients and trials.

stereoelectroencephalography · sensory mode decoding · event timing · high gamma · number cognition · label permutation

Admitted 1 citing publication

References

5
  1. usesA fixed 70–110 Hz decoder does not detect sensory mode transfer across four formats in 12 SEEG recordings sci:sha256:ec2ade151571af9d4088a48b2e6e6aa24be6dcbe46c6b087cc0f601c44f62770

    We used its fixed trial cohort, forward split, 70–110 Hz feature code, and ridge classifier. We rebuilt all published-clock arrays for the four selected patients and obtained exact matches. The full source package is preserved.

  2. usesJoint label changes preserved shared task labels and class counts in 944 SEEG trials from 12 patients sci:sha256:717ef2758863aad2f2221a8aa82783553e26746ebcb3c4ad60f706c6d0ecb842

    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.

  3. usesConstant event time shifts aligned most light sensor steps in four ds005415 v1.0.0 SEEG recordings with nominal sample timing sci:sha256:77c82896c1a121debd08d946593404b861ef8b532c494f99a24d83ab5c7c1c5b

    We used its four sensor-supported constant event origin shifts and nominal sample timing assumptions. We applied the shifts before neural feature extraction. The full source package is preserved.

  4. usesdoi:10.18112/openneuro.ds005415.v1.0.0

    The analysis used SEEG data from OpenNeuro ds005415 version 1.0.0.

  5. doi:10.1371/journal.pone.0313155

Cited by

1 publication

Public catalog entries on Sciscape.