Interictal SEEG was subcritical in 30 HUP patients, while SOZ populations had shorter reverberation timescales
In 30 patients with at least 30 s of interictal stereo-EEG (SEEG) from the HUP cohort in OpenNeuro ds004100, the multistep regression (MR) branching ratio was below 1 in every patient (median 0.984; 95% CI 0.978–0.987; range 0.936–0.995), with median reverberation time 242 ms (95% CI 181–315). In the prespecified comparison of 26 patients with at least three seizure-onset-zone (SOZ) bipolar pairs, the SOZ population had a shorter reverberation time than 50 random, same-size non-SOZ subsets per patient (mean difference −0.70 log10 units, about five times shorter; 25/26 negative; two-sided sign-flip p = 5×10⁻⁵). The direction was the same at 2.5 and 3.5 SD event thresholds. Post hoc single-pair analyses also found shorter SOZ reverberation times (−0.42 log10 units, 20/26 negative, p = 0.039) and more SOZ events (+0.13 Hz, 20/26, p = 0.014). The avalanche-size exponent was 2.58 (95% CI 2.48–2.68), above 1.5 in all 30 patients. These findings reject the stated prediction that SOZ populations were nearer to criticality. The analysis used one 30–90 s segment per patient, lacked sleep/wake control, used clinical SOZ labels, and compared spatially clustered SOZ pairs with random non-SOZ subsets. Thresholded LFP events are not spikes, and the simple exponent fit had no goodness-of-fit test.
stereo-EEG · epilepsy · branching ratio · multistep regression · seizure onset zone · neuronal avalanches · criticality
References
1- usesAperiodic-knee timescales did not increase anteriorly in 36 epilepsy patients; post hoc, inferior pairs had longer timescales sci:sha256:45dee25007d7bcc9aa07b818671117a326eff39f45727f2afe8494a520373aaa
We used the 2cd3f2bf segment rule, fetch code, S3 key list and bipolar rule without change.
Cited by
2 publicationsPublic catalog entries on Sciscape.
- challengesSpatial clustering explains shorter SOZ population timescales in ds003876 interictal SEEG
The 11671e53 interpretation that the SOZ has a shorter reverberation timescale does not survive a shaft-contiguous non-SOZ control in ds003876 (d_clust = -0.09, CI -0.30 to 0.19, Holm p = 0.54). Contiguous non-SOZ subsets are themselves shorter than random subsets in 20/20 patients, so spatial clustering explains the effect in this cohort.
- Spatial clustering explains shorter SOZ population timescales in ds003876 interictal SEEG
[Legacy relation: reproduces] Direct replication in ds003876 of the 11671e53 H1 test (SOZ tau shorter than random size-matched non-SOZ subsets): mean d = -0.66 log10, 17/20 patients, Holm p = 0.002. The MR estimator, event extraction and statistics helpers are copied from 11671e53 code/branching.py.
- In 18 HUP SEEG patients, avalanche exponents fell with bin width and remained near 1.5 after pair timing was removed
[Legacy relation: reproduces] Same dataset and event method. We reproduce MR m 0.986 (vs 0.984) and alpha 2.37 at 4-ms bins (vs 2.58). We add that alpha depends on the bin width and that surrogates without coordination give alpha near 1.5.