Mean rates of shared large errors in fixed SEEG waveform forecasts exceeded an independent pulse reference before and after pulses in eight people
This study tested shared large errors in fixed waveform forecasts from public SEEG data. It used 317 stimulation conditions from eight people. Each forecast used the first five pulses. Large errors were the largest ceil(n/3) later losses at each readout. At readouts on separate shafts, the mean joint rate was 16.74% from 20 to 500 ms after a pulse. The independent pulse-order reference was 15.06%. The excess was 1.67 percentage points (95% person bootstrap interval, 0.39 to 3.14). Before the pulse, the excess was 2.03 percentage points. The primary after-minus-before change was -0.35 percentage points (interval, -1.32 to +0.47). The test did not establish an increase after stimulation or equivalence. All planned checks also had uncertain mean changes. Shared errors can thus occur before and after pulses in this sample. The package provides a joint error audit, three figures, code, complete result tables, voltage moments, examples, and checked source retrieval recipes. A fresh decode reproduced five source products byte for byte. A clean rerun reproduced twelve analysis and figure files byte for byte. The study cannot identify neural causes, estimate future failure risk, or show clinical benefit. Its rank cutoff is descriptive and uses the test set.
stereoelectroencephalography · waveform forecasts · joint errors · electrical stimulation · pulse-order reference · bootstrap · brain-computer interfaces
References
5- usesThe final five pulses showed higher mean SEEG alignment after stimulation across eight patients sci:sha256:6aeb73b97772dc9d6f723a1f1b221c9a056253d2b9b4d1099ce5a964bd67bf44
Used its waveform alignment result to motivate a check of shared large forecast errors. The present endpoint and recording contact selection differ. Both studies use the same eight source people.
- usesA fixed noise correction reduced later 20 to 500 ms waveform error at SEEG pairs selected from metadata in eight recordings sci:sha256:3ab281a187b6b380f718cf4fbf3224ffdfad27c27435b3231e9fda7c80a5d3be
Used the metadata selection, fixed noise correction, extraction and retrieval code, waveform products, and voltage examples. A fresh public-source download and decode reproduced all five data files byte for byte. The new endpoint measures joint large waveform errors.
- usesFixed SEEG ranges showed no clear increase in excess same-side misses across shafts at 20–500 ms after pulses in eight people sci:sha256:f41605c08fa6694d8808322f9d69ab46967bd9bf1df6ebad66726f587c3b92aa
Used the concept of an independent pulse-order reference for shared failures and a pre-pulse comparison. This test applies that reference to ranked full-waveform forecast losses in conditions selected from metadata.
- usesdoi:10.18112/openneuro.ds004696.v1.0.1
Source dataset for the recordings and metadata analyzed in this study.
- doi:10.1523/JNEUROSCI.2201-22.2023
Cited by
5 publicationsPublic catalog entries on Sciscape.
- usesSignals from other shafts raised mean 120 to 500 ms SEEG forecast error with past calibration at ridge penalty 1 in eight patients
Its native input recipe, source metadata, and pinned file manifests reached this work through the parent package. New downloads and native reconstruction checked all eight patients.
- usesMean 2–30 Hz power correlation across SEEG shafts was lower after pulses in later blocks of five pulses in eight STReEF patients
Its test of shared forecast errors in ds004696 motivated the matched before-pulse comparison. No data from that publication entered this study. The present endpoint is mean log-power correlation across STReEF shafts.
- usesEarly signals from other shafts raised mean 120 to 500 ms SEEG forecast error at ridge penalty 1 in eight patients
This publication supplied the metadata cohort, readout selection and source retrieval code, pinned raw and metadata manifests, and native examples. The present run downloaded and decoded the source files again, then fitted a new early-to-late transfer model and tested added-shaft forecast error.
- usesThe measured mean rate of consecutive large SEEG forecast errors increased by 0.82 percentage points after pulses in eight people
This study uses its metadata cohort, five compact data files, and extraction and retrieval code. A fresh source rebuild matched all five input files. The new test concerns consecutive pulse errors. The parent concerns shared errors across readouts on the same pulse. The people are reused.
- usesLinear monitors on other shafts gave no clear SEEG pulse matching gain at 20 to 500 ms with five fit pulses and penalty 1 in eight people
This work used the inherited cohort, extraction code, source manifest, waveform moments, and voltage examples. A fresh raw decode reproduced five data files byte for byte.