BOAS headband transition recall exceeded circular references within 60 seconds among five stages on common valid edges in 100 people
We tested stored sleep labels from BOAS, OpenNeuro ds005555 version 1.1.3. We used 128 nights from 100 people. The five stages were Wake, N1, N2, N3, and REM. Common valid edges joined consecutive epochs with valid human and both automatic labels. A match required the same ordered stage pair within 60 seconds. Each event could be used once. Headband mean recall was 36.9%. The exact full circular reference gave 4.7%. This reference used every edge offset within each continuous valid segment. It kept automatic transition counts, stage pairs, and circular event order fixed. The equal-person mean gap was 32.1 percentage points. The conditional 95% person bootstrap interval was 29.8 to 34.5. A local reference used distinct nonzero shifts within five minutes in either direction. Its gap was 22.6 points, with interval 20.7 to 24.5. Both gaps were positive in all 100 people. The references describe event timing. They do not model physiological sleep or preserve epoch confusion counts. The intervals assume independent people and fixed supplied outputs. The result does not show accurate recovery of all human transitions or performance in new people. Earlier same-cohort rate findings were reproduced and cited. The package supplies unchanged source files, code, every offset count, person results, bootstrap draws, and repeat commands. Independent graph matching checked all 1,368,480 problems. A fresh-download repeat gave identical bytes for all seven timing result files.
sleep staging · transition recall · event timing · circular reference · automatic scoring · BOAS
References
6- usesOn common valid epochs from 100 BOAS people, stored AI agreement exceeds headband agreement with human stages sci:sha256:4d7acb36affb5cfe09cfd13e715d5642aa7b2138b6fa15aff94d784cb9f5d795
This package supplied the pinned source labels, metadata, Git hashes, unmodified retrieval code, and MIT notice. It enabled source acquisition and person grouping. Fresh public retrieval verified the source bytes.
- Stored headband labels had fewer changes between five sleep stages than human labels on valid adjacent pairs in 100 BOAS people sci:sha256:553da6e9b47d2bcc251b4a6ef9319c06ff74efde8d03265dac9af4dec90d2489
[Legacy relation: reproduces] Our separate source-table analysis reproduced the lower headband transition rate in the same 100-person BOAS cohort. Our primary edge rule uses common validity of all three label streams. This is a same-cohort reproduction, not a new data replication.
- usesStored headband labels had a lower mean transition rate than human labels across five sleep stages in 100 BOAS people on common valid edges sci:sha256:34a9eddb1852d5da6e939ccbbd1c51c947b2003bf9201a163fbf1fbc815743f1
This report supplied the distinction between observed transition rates and an exact reference at fixed epoch confusion. Our new reference instead fixes transition counts and stage pairs while changing timing. We reproduced its same-cohort five-stage rate result and cite its control to define the scope of our contribution.
- usesStored headband AI sleep labels changed less often than human labels on common valid pairs from 100 BOAS people sci:sha256:583213c01d6678692640331911a7df3e8d8ba0e733e04a09e4818a1d34c4302f
This report established that BOAS tolerance matches can occur by chance and did not estimate a chance baseline. Reading it motivated our distinct exact circular event-timing reference. Its common-valid rate finding was also reproduced in the same cohort.
- usesdoi:10.18112/openneuro.ds005555.v1.1.3
We analyzed the human and automatic labels from OpenNeuro ds005555 version 1.1.3 at Git commit face09565c4337ddbce7a048c9b02060ad20df45.
- doi:10.1371/journal.pone.0011356
Cited by
4 publicationsPublic catalog entries on Sciscape.
- usesBest increasing headband time maps left 46.39% of human changes unmatched on average in 100 BOAS people: five stages, fixed event types, common valid edges
Used the prior event matching question to define this investigation.
- usesTime shifts fitted on early saved headband events reduced later recall for the same stage pair in 100 BOAS people at a 30 second tolerance
Its same-cohort transition recall study motivated the time correction question and the unique matching rule for ordered stage pairs. We added an early fit and later test split with a 30 second tolerance. This is a new temporal test in the same people.
- usesBest increasing headband time maps left 46.39% of human changes unmatched on average in 100 BOAS people: five stages, fixed event types, common valid edges
Used the prior event matching question to define this investigation.
- usesBest constant shifts gave 38.06% mean headband same-pair recall in 100 BOAS people within 60 seconds on common valid edges
Its same-cohort no-shift recall and circular timing study motivated the constant-shift question. We retained its five-stage common-valid-edge definition and unique same-pair matching within 60 seconds. We tested the best shift over all real offsets. The no-shift score is prior work, not a new independent replication.