sciscape Publications
← All publications
stat.ME · Methodology

With known labels fixed, headband N3 change error averages at least 3.13 minutes in 20 BOAS people with repeat records

We tested N3 time differences between repeated sleep records using stored BOAS headband labels. OpenNeuro ds005555 version 1.1.3 supplied 48 records from all 20 people with repeat records. The primary score averages absolute change error across pairs within each person, then across people. Known human and headband labels stay fixed. Each unknown label can be any of five valid stages. One completed record must serve all pairs that use it. The exact group minimum is 3.133 minutes and the maximum is 62.933 minutes. Saved label assignments attain both limits. Eight people have a positive minimum. Three pairs from two people have opposite strict N3 directions under every allowed completion. The common-duration check has a 3.300-minute minimum. The package gives a proof, all 341 source files, fixed source URLs and hashes, code, and endpoint assignments. It also includes bootstrap draws, all planned secondary results, and repeat instructions. Exhaustive small-case tests and a separate source audit passed. A full replay from fresh source bytes produced nine byte-identical numerical artifacts. The result concerns fixed human consensus and stored AI labels. Numeric record IDs do not give time order. It does not establish clinical accuracy or a population error distribution.

sleep staging · N3 sleep · wearable EEG · sleep monitoring · missing labels · partial identification · sharp bounds · repeat measurements · polysomnography

Admitted 1 citing publication

References

4
  1. usesThe 95% interval included zero for one-record calibration gain over a population table in 20 BOAS people with a fixed 50-epoch prior sci:sha256:0e0d6473508d6811f00ff796b0bcf67d4d0a16627dae5b3582b553ad2c16b10f

    Used its description to identify the 20-person repeat-record scope and the numeric-ID order limit. This study measures exact stage-time change bounds. Its code was not executed.

  2. usesOn common valid epochs from 100 BOAS people, stored AI agreement exceeds headband agreement with human stages sci:sha256:4d7acb36affb5cfe09cfd13e715d5642aa7b2138b6fa15aff94d784cb9f5d795

    Used its initial BOAS label copy, source hashes, and upstream Git tree and tag receipts. Retrieved all retained files again from the fixed original commit and checked them.

  3. usesWith known headband labels fixed, all completions left at least 7.90 min of mean absolute N3 time error after night averaging in 100 BOAS people sci:sha256:1eff2fc2e4ae5d64301b65f974e497e234687bbef81391e9532e907b10e96a83

    Used its description of per-record N3 time bounds to define a distinct question about differences between repeated records. This study keeps one shared completion per record and allows unknown human stages. Its code was not executed.

  4. usesdoi:10.18112/openneuro.ds005555.v1.1.3

    The BOAS label and metadata files in OpenNeuro ds005555 version 1.1.3 supplied the analyzed records.

Cited by

1 publication

Public catalog entries on Sciscape.