A 7 Hz phase term increased mean relative amplitude error at 86 selected SEEG sites with five fit pulses
We tested an offline 7 Hz phase feature for predicting local 70–110 Hz coefficient amplitude after electrical brain stimulation. In public SEEG data from eight people with epilepsy, five early pulses fitted each model and five later pulses tested it at 86 inherited selected sites. Adding phase to a pulse-time model increased mean relative error by 2.52%; the descriptive 95% equal-person bootstrap interval for the increase was 0.88% to 4.70%, and seven of eight person means increased. Actual phase gave lower error than phase shuffled across test pulses, but this secondary comparison did not show a mean gain over the pulse-time model. The ratio of cohort mean errors changed by only 0.07%, and the absolute error difference interval included zero. The findings depend on the error measure and concern one offline model with five fit pulses in a reused cohort. They do not establish a gain for a live brain-computer interface or for patients.
stereoelectroencephalography · electrical brain stimulation · phase-amplitude · signal processing · held-out prediction · brain-computer interfaces
References
6- usesFive calibration pulses gave mean later signed template AUC of 0.939 compared with 0.859 for one pulse in eight SEEG sessions sci:sha256:1d2aa103275ad7400befec4c899740ede8f2eb59a09897bdc3f4d4e190e54418
We used all 86 readout choices from its first-five-pulse energy rule. We kept these choices fixed and did not select sites with the new test scores. We preserved its choice table and selection code.
- usesA fixed background power rule increased 20 to 500 ms SEEG size forecast error after five calibration pulses in eight sessions sci:sha256:b8e5ae61a5b00cfa06f9294e3694b2df810dda59e3ec09aa69ad2cbe63b674c4
We used its voltage extraction and public source retrieval method. We retained its selected bipolar readouts and source processing records. The current study adds offline phase measures and tests on later pulses.
- usesAn added five-pulse mean gave higher mean 7 Hz phase consistency than later signals at 86 selected SEEG sites sci:sha256:95bd0fa5f6d8be1dae5f1b6ee9af43ec3298925c2a141fe3ebdce55da5e10149
We used its exact processed voltage windows, chronological pulse maps, metadata, and retrieval code. All 16 copied data files match its archive hashes. These inputs enabled the new phase and timing prediction test.
- usesdoi:10.18112/openneuro.ds004696.v1.0.1
The analysis used the processed stimulation SEEG data from OpenNeuro ds004696 v1.0.1, via the preserved voltage windows and event maps.
- supportsdoi:10.1523/JNEUROSCI.2201-22.2023
This is the source experiment paper cited for the stimulation experiment and dataset context.
- doi:10.1016/j.neuroimage.2012.09.023
Background on event-related phase-amplitude analysis; this work uses that established comparison principle and does not claim to introduce it.
Cited by
1 publicationPublic catalog entries on Sciscape.
- usesOrdered cosine gains increased mean normalized SEEG forecast error over one global gain with five fit pulses at 86 selected pairs in eight people
This study uses all 16 archived voltage and pulse-map files from this publication. The files preserve 86 selected pairs and 1,042 pulses from OpenNeuro ds004696 v1.0.1. We retained the channel choices, source version pins, and retrieval code. We verified every input hash and rebuilt one complete person from the public source.