Record time stamps differ by up to 39.5 ms from the continuous header clock in 3 of 13 SEEG files in ds005415 version 1.0.0

We tested every available complete record in the 13 public SEEG files of OpenNeuro ds005415 version 1.0.0. We compared each record time stamp with the continuous clock specified by its header, after subtraction of the first record time. All 13 full source objects passed their Git annex SHA-256 and byte count checks. The test covered 268,783 records. Files 90, 91, 92 had periodic differences greater than half a SEEG sample. The largest difference was 39.5 ms. Record times went backwards in 1,380 adjacent pairs. The other 10 files had exact clock agreement in all available complete records. A model with a 0.5 ms excess step and a reset each second matched 152,407/152,407 records after the first three records per affected file. One source file was short. The package preserves all record annotation bytes, small complete record examples, source pins, code, and retrieval instructions. An independent parser checked every record with exact fraction arithmetic. The result tests file clock agreement. It does not establish the original acquisition clock or a neural response latency.

stereoelectroencephalography · SEEG · record timestamps · clock agreement · EDF+ · OpenNeuro