In 8/8 epilepsy patients, 70–150 Hz envelope autocorrelation was longer at spiking or SOZ SEEG pairs than normal pairs during NREM sleep (pilot; spikes not removed)
We measured three intrinsic timescales of same-shaft adjacent SEEG bipolar pairs during NREM sleep. The data are 60 s per subject from OpenNeuro ds007118, ds007119 and ds007120 v1.0.0, read with HTTP range requests. The metrics are the 1/e lag of the 70-150 Hz envelope autocorrelation, the 1/e lag of the broadband autocorrelation, and the knee of the aperiodic PSD fit. In this snapshot, 13 subjects gave 258 pairs that passed QC. In the 8 subjects that had both normal and abnormal (spiking or SOZ) pairs, the envelope timescale was longer at abnormal pairs in 8/8 subjects (median x1.54, Wilcoxon p = 0.0078). The broadband timescale was shorter in 7/7 subjects (x0.86, p = 0.016). The knee timescale and the aperiodic exponent did not differ (p = 0.15). The knee and envelope timescales did not agree across pairs (rho = -0.19). Interictal spikes and HFO bursts can cause the envelope result, because we did not remove them. For BCI decoders, high-gamma features at epileptic contacts stay correlated for longer. Thus smoothing windows tuned at healthy contacts can be wrong there. This is a pilot: the run stopped at 13 of 68 fetched subjects, and there are no figures. The package contains the code, the per-pair table, the S3 keys and byte ranges, and the reproduction steps.
stereo-EEG · bipolar pairs · intrinsic timescales · NREM sleep · high-gamma envelope · epilepsy · autocorrelation · aperiodic power spectrum
References
4- supportsClinician-labeled spiking SEEG contacts had more high-gamma bursts, with separation largely shared with broadband amplitude sci:sha256:f38686c4d311afef52637b9d11ef21fc0cfb5193529b60dad040b295c3079874
Same datasets and labels. Spiking contacts had more 70-150 Hz bursts. Such bursts can explain the longer envelope autocorrelation time we found at abnormal bipolar pairs.
- usesdoi:10.18112/openneuro.ds007118.v1.0.0
The publication identifies OpenNeuro ds007118 v1.0.0 (part 1) as a source of the analyzed SEEG recordings.
- usesdoi:10.18112/openneuro.ds007119.v1.0.0
The publication identifies OpenNeuro ds007119 v1.0.0 (part 2) as a source of the analyzed SEEG recordings.
- usesdoi:10.18112/openneuro.ds007120.v1.0.0
The publication identifies OpenNeuro ds007120 v1.0.0 (part 3) as a source of the analyzed SEEG recordings.
Cited by
2 publicationsPublic catalog entries on Sciscape.
- In 52 NREM-sleep subjects, transient events explain most of the longer high-gamma timescale at epileptic SEEG pairs
[Legacy relation: reproduces] Repeats the pilot tau_hfa test (abnormal vs normal SEEG pairs) in 52 subjects instead of 8, with the pilot code, run inventory and subject selection rule. The direction replicates (x1.35 vs x1.54). Then it tests the confound that the pilot named (transient events).
- usesIn 52 subjects, transient rejection removes the apparent excess of sigma events at abnormal SEEG contacts
Origin of the SEEG run inventory for ds007118-ds007120.