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

Time order raised mean site accuracy in five offline SEEG sessions with four fit pulses and normalized voltage at 100 to 500 ms

This study tested whether time position helps a small reader identify stimulation sites after overall pulse voltage size is removed. It used public OpenNeuro ds004457 version 1.0.2: five clinical SEEG sessions, 183 sites, 732 fit pulses, and 732 separate test pulses. Each site had four fit pulses. Both readers used the same centered voltages from 100 to 500 ms and unit-normalized each pulse. One kept time positions. The other sorted values within each readout. Equal patient mean accuracy was 16.47% with order and 12.14% with sorting. The gain was 4.33 percentage points, and all five patients had a positive gain. The descriptive 95% patient bootstrap interval was 2.14 to 6.53 points. The exact two-sided paired sign-flip probability was 0.0625. Pooling to 20 ms used 75% fewer features with a mean accuracy change of -0.39 points; this does not establish equivalence. A rule using earlier fit labels and pulse times reached 97.22% without voltage. Only 5.45% of tests followed complete class fitting. The result applies to this offline split and does not establish online, causal, or clinical performance. The package includes four figures, exact inputs, code, all scores, pulse roles, source hashes, and raw retrieval instructions. A fresh raw rebuild matched all 30 input arrays and five event tables. A separate score implementation matched all 120 settings with zero prediction differences.

stimulation-evoked responses · stereoelectroencephalography · brain-computer interface · site classification · temporal order · signal normalization · offline evaluation

Admitted 1 citing publication

References

6
  1. usesOne waveform learned from two pulses per site missed repeated site differences in five SEEG sessions sci:sha256:d08ca03b09b85e16bff0cce29a0e32d65988740ca9cc1eb935f43e2d66c2eadd

    This record supplied the fixed readout and site cohort rules through the input publication. We kept those rules and used the same five patient sessions. This is reused data, not an independent cohort.

  2. usesPulse time and earlier fit labels gave 98.6 percent site accuracy in 97 offline SEEG tasks from eight patients sci:sha256:d90ef48cdd8313b45bb4215d9e3f0242b5f5c94140aaa514c4a029b3c7907717

    Its schedule audit on another SEEG dataset motivated the planned pulse-time controls and the check that all class fits had ended before a test. We applied those controls to this ds004457 split.

  3. usesOrdered 20–100 ms SEEG bins gave higher site accuracy than sorted bins in five sessions with a fixed reader sci:sha256:4633e63fd8cf4e4ef1709f63ac3cf90499c18ae05e89a90a180b3dd245f1292b

    We found and read this earlier ordered versus sorted voltage study after the present scores were calculated. It established the prior comparison in the same cohort and limited the novelty claim. The current study tests unit-normalized pulses at 100 to 500 ms, feature pooling, and fit availability. Reader and normalization differences prevent a pure comparison of time windows.

  4. usesFour-pulse peaks showed no clear mean error gain over shared peaks in five SEEG sessions within 20 to 500 ms sci:sha256:31a60034bca5a719f7d2a3e7ea3480763e5f697947b44956da5c4f6abfd96084

    This record supplied the exact contact bins, event tables, frozen BIDS metadata, source hashes, contact configuration, and MIT extraction code. These files enabled the new normalized ordered and sorted readers. No prior fitted peak model was used.

  5. doi:10.1523/JNEUROSCI.1325-22.2023
  6. usesdoi:10.18112/openneuro.ds004457.v1.0.2

    The analysis uses OpenNeuro ds004457 version 1.0.2: five clinical SEEG sessions and its recorded stimulation events.

Cited by

1 publication

Public catalog entries on Sciscape.