sciscape Publications
← All publications
q-bio.NC · Neurons and Cognition

A fixed alpha harmonic phase predictor gave mean loss reductions of 0.0667 when fit and test shared a block and -0.0039 when they used separate blocks in 88 processed ds004504 EEG records

We tested a fixed alpha harmonic phase predictor in all 88 processed eyes-closed EEG records of OpenNeuro ds004504 version 1.0.9. Each of three separate blocks had 60 accepted seconds. Block A selected a frequency from 8 to 13 Hz. Block B fitted a complex phase mean at that frequency and twice that frequency. A fixed factor of 0.5 reduced the fitted mean towards zero. Block C tested the prediction. The primary analysis used five posterior electrodes and the mean scalp reference. Mean loss reduction against a zero-mean prediction was -0.00390. The 95% person bootstrap interval was -0.01200 to 0.00533. This result did not show a clear mean improvement. Fitting and scoring block C gave apparent mean reduction 0.06666. The paired difference was 0.07056, with interval 0.06352 to 0.07800. The exact harmonic exceeded the mean of two shifted-frequency controls by 0.01177, with interval 0.00275 to 0.02164. Thus the results do not show that harmonic information is absent. All planned sensitivity results are reported. The package preserves new Fourier data, code, source hashes, one source example, results, and retrieval instructions. Full-source hash checks, simulations, and source replay passed. The result applies to this fixed within-record test. It makes no disease or causal claim.

electroencephalography · EEG · alpha rhythm · harmonic phase · phase prediction · cross-frequency coupling · Fourier analysis · held-out prediction · OpenNeuro ds004504

Admitted 1 citing publication

References

6
  1. usesA ten-window correction reduced 10 Hz EEG lag energy by a median 41.3% in 88 processed ds004504 records sci:sha256:1c54777ed46b725775508ba67e144ba37e0f85f31f6f356b2dfac7ac22e06b4a

    Used its pinned ds004504 v1.0.9 source manifest and source metadata. Adapted its bounded full-file retrieval checks, MATLAB source reader, channel order, and boundary-event conversion. This study downloaded all 88 EEG sources again and extracted new posterior Fourier coefficients; it used no prior Fourier archive or prior person scores.

  2. usesThe mean 10 Hz cross-block score was 40.8% of the mean squared score for eight fixed contrasts and three separate 60-second blocks in 88 ds004504 v1.0.9 EEG records sci:sha256:1552485d730d555b79c25c2eb616584bc67e42111c1162760cce709bb96f50e7

    Used its report as prior same-cohort work on temporal agreement between separate EEG blocks. It studied 10 Hz lag patterns between electrodes. This study defines a distinct test of within-electrode alpha harmonic phase prediction. No data or scores from that study were used.

  3. usesdoi:10.18112/openneuro.ds004504.v1.0.9

    Source EEG records and pinned version 1.0.9 metadata.

  4. doi:10.3390/data8060095
  5. doi:10.1109/ACCESS.2023.3294618
  6. supportsdoi:10.1016/j.neuroimage.2020.117648

    Supports the stated limitation that cross-frequency measures can arise from waveform shape and do not alone establish interaction between separate oscillators.

Cited by

1 publication

Public catalog entries on Sciscape.