In 8 epilepsy patients, clinical SEEG band-power LDA decoded key side in 1 and movement vs fixation above AUC 0.9 in 3 of 7
We analyzed SEEG band power from 8 patients performing a forced two-choice arrow-key task, using clinically placed contacts without motor-cortex targeting. Bipolar-pair features and a shrinkage LDA with 5-fold cross-validation decoded the left or right key on correct trials. Only 1 of 8 patients showed above-chance pre-press decoding (AUC 0.83; permutation p=0.02); group mean AUC was 0.53 (Wilcoxon p=0.42). An early control window showed a small group effect (mean AUC 0.54, p=0.012), but it was not a clean pre-cue control and is not interpreted as movement information. On 33 error trials, the decoder followed the pressed key in 14 (binomial p=0.49), which did not separate motor from stimulus coding. As a positive control, movement versus fixation was decoded with AUC above 0.9 in 3 of 7 valid patients. The results suggest that these clinical SEEG contacts often showed whether movement occurred but seldom showed its side. They apply to this task, these band-power features, a linear decoder, and 50-permutation tests. The sample was small, motor-cortex coverage varied, error-trial power was low, and the pre-move window followed the go cue.
SEEG · stereo-EEG · brain-computer interface · movement decoding · band power · linear discriminant analysis · OpenNeuro
References
2- usesdoi:10.18112/openneuro.ds004473.v1.0.2
The analysis used the OpenNeuro ds004473 snapshot identified in the source report as version 1.0.2.
- doi:10.1088/1741-2552%2Facae0a
Cited by
4 publicationsPublic catalog entries on Sciscape.
- usesIn 8 patients, a 100 Hz limit changed high-gamma-only SEEG detection but had little effect on multiband AUC
Same dataset and movement detection question; we add a test of the sensing band limit.
- supportsA detector trained on one arrow key detects the other key in SEEG, with non-inferior 70–150 Hz press-vs-rest decoding in eight subjects
Our move-window key-side decoding (median AUC 0.54, only sub-2 high) agrees with its result that band power decodes the key side in only 1 of 8 patients. Our cross-side transfer result gives a reason: the key-press signal is mostly common to both keys.
- supportsIn two SEEG patients, a band-power LDA separated key presses from rest but also detected fixation before the go cue
Same dataset. Our press against fixation AUC (0.87 in sub-1 and sub-2) agrees with its result that SEEG band power shows that a movement occurs. We add detection timing and a fixation-period control.
- usesIn 8 SEEG patients, ERP covariance decoders had lower AUC than voltage decoders, while high-gamma covariance added no gain
It is a prior band-power LDA analysis of left vs right key press in the same 8 ds004473 patients. Our high-gamma log-variance and covariance decoders test related band-power features.