Personal alpha spatial filters showed no clear mean balanced accuracy gain over personal mean alpha power with an average reference one month after calibration in 60 ds004148 adults
We tested a fixed personal alpha spatial filter in OpenNeuro ds004148 version 1.0.1. Each state supplied a 60-second EEG crop. Session 1 set each person’s weights and thresholds. Session 3 tested the same rules about one month later. Both methods used the same eight channels, average reference, and accepted windows. In 60 adults, mean balanced accuracy was 89.68% for the spatial method and 88.66% for mean power. The paired gain was 1.01 percentage points. Its 95% person bootstrap interval was −0.38 to 2.39 points. The interval included zero. The test gave no clear mean gain; it does not show equal performance. The package preserves new covariance data, fitted rules, predictions, source versions, code, and checks. Fixed state order and eye-movement component removal in open-eye source files limit interpretation.
electroencephalography · alpha rhythm · spatial filter · balanced accuracy · test-retest reliability · eye state · personal calibration
References
6- usesPerson thresholds from 60 seconds per state increased one-month eye-state balanced accuracy by 7.6 percentage points in 30 ds004148 adults using offline 8-13 Hz EEG spectra sci:sha256:7d2065009c4dfcc307d2bd9a1322334ba4d05630f6f2c88cbdfbfd22825b2d1d
This publication supplied evidence about personal power thresholds across a month in ds004148. That result helped define this question and the personal mean-power baseline.
- Personal alpha spatial filters improved run 2 balanced accuracy over mean alpha power in 42 ds007987 adults after calibration within one visit sci:sha256:9713af39d1278baa2b5b7f2f53c3cfa51fc29817b6ddd51a09dec3ed080ab68d
[Legacy relation: reproduces] This publication supplied the fixed spatial filter, power baseline, and earlier within-visit result. We repeated the method in a separate cohort across a month. The original method code is preserved for credit and inspection.
- usessci:sha256:eb6da8fdd843c3b66dd19edef3df5eb85f888448f106108e611352a1c6388159 not known to Sciscape
This publication supplied source headers, sidecars, full-file hashes, and pinned S3 versions for ds004148. These records enabled new raw-crop retrieval and covariance extraction. We checked all source metadata against Git and checked two complete file hashes again.
- usesdoi:10.18112/openneuro.ds004148.v1.0.1
Source dataset analyzed in this study: OpenNeuro ds004148 version 1.0.1.
- supportsdoi:10.1038/s41597-022-01607-9
Provides the cohort and release-method details cited for ds004148, including the state-specific processing described in the report.
- doi:10.1016/j.neuroimage.2021.118809
Cited by
4 publicationsPublic catalog entries on Sciscape.
- usesPosterior channel differences did not show higher alpha reactivity agreement across one month in 60 ds004148 adults with fixed windows
This study extracted the ds004148 covariance features reused through the direct input archive. We retained its channels, crop, filters, units, and covariance definition. We did not use its personal fitted filters or state predictions.
- usesPersonal alpha spatial filters had no clear mean balanced eye-state accuracy advantage over peer filters after one month in 60 adults with personal thresholds
We used all 360 alpha covariance records, source metadata, fixed source versions, crop hashes, and extraction code. These files enabled the month test and two fresh signal replays.
- usesPosterior alpha covariance showed person links after total power removal across one month in 60 ds004148 adults with closed eyes
Supplied all 360 alpha covariance records, the raw example, source metadata, extraction code, and pinned raw crop ledger. The new analysis tests normalized spatial person links instead of the source eye-state classifier.
- usesFixed zero-sum alpha filters with the lowest session 1 ratio had a positive mean session 3 contrast in 60 ds004148 adults
This study uses all 360 saved covariance files, the pinned crop ledger, source metadata, one raw crop, and the original extractor. It tests minimum-ratio filters with zero-sum weights and later log power ratios. It uses the same cohort.