Sets of eight sites selected with four calibration pulses per site gave 97.5% mean SEEG site accuracy on later pulses in five sessions
We tested whether four calibration pulses at each candidate site can select eight stimulation sites with distinct later SEEG responses. The data had 183 candidate sites in five patient sessions. Selected sets gave 97.5% mean site accuracy on four later pulses per selected site. Random sets gave 60.9%. Sets chosen for large responses gave 93.8%. Calibration and test pulses were separate. The before-pulse control gave 12.5%, equal to chance. The method required calibration at all candidate sites. Five patients are too few to establish a population effect. The test does not show performance on another day or useful brain control. The package has derived data, figures, source hashes, code, and reproduction instructions.
SEEG · electrical stimulation · site selection · response classification · calibration · brain-computer interface · within-session validation
References
4- usesA fixed covariance rule gave higher SEEG site accuracy than a diagonal rule in offline tests from five patient sessions sci:sha256:1407cd9ac712a4bcd551a3738cc90c3fab309b871572cbd6fc644b41555f7886
This package supplied the voltage bins, fixed recording targets, pulse split, extraction code, cohort reader, and fixed half-covariance decoder. We checked the source hashes and rebuilt all inputs from the raw recordings. We added calibration-based site selection and tested separate later pulses.
- usesOne waveform learned from two pulses per site missed repeated site differences in five SEEG sessions sci:sha256:d08ca03b09b85e16bff0cce29a0e32d65988740ca9cc1eb935f43e2d66c2eadd
Its test showed repeated site differences outside one learned waveform direction in this cohort. That result helped us choose a fixed-size site selection question. We did not copy data or code from that package.
- usesdoi:10.18112/openneuro.ds004457.v1.0.2
The analysis used public SEEG data from OpenNeuro ds004457 version 1.0.2.
- doi:10.1523/JNEUROSCI.1325-22.2023
Cited by
4 publicationsPublic catalog entries on Sciscape.
- usesMinimum split selection gave 40.0% mean SEEG accuracy at 100–500 ms for eight sites in five offline sessions with four calibration pulses per site
Its greedy selector used a score from two calibration halves. We extended that selector to all three balanced splits of four pulses. We also kept its first split rule as a fixed comparator.
- usesEarly and delayed selection gave 42.5% and 45.6% mean delayed SEEG accuracy for eight remote sites per session in five 6 mA sessions
Its voltage bins, event tables, fixed recording targets, source metadata, retrieval and extraction code, calibration metric, and greedy rule enabled the new response-window site-selection test. We reproduced its original selected sites, pulse indices, all 160 primary predictions, and 97.5% mean accuracy exactly. We then applied stricter remote-shaft and 6 mA rules. The new task uses the same source patients and is not an independent patient replication.
- usesTwo calibration pulses gave a mean score gain of 0.062 bits over eight sites for 20 to 100 ms SEEG responses in five sessions
This package supplied the compact voltage inputs, source metadata, pinned file manifest, extraction code, and contact eligibility rules. We rebuilt its inputs from the raw recordings and added two separate pulses to choose the site count.
- usesExchanged later voltage changed a mean 43.1% of site labels before pulses in five SEEG sessions with a centered 30 Hz FIR filter
This work supplies the fixed recording targets, metadata rules, snapshot metadata, pinned raw-file hashes, and retrieval code. The present run retrieves the native MEF files and extracts a new all-eligible-site cohort. It does not use the source selected eight-site sets or its derived after-pulse feature arrays.