The 30–45 Hz SEEG aperiodic exponent did not flatten with age in 60 adults with epilepsy

We tested whether the age-related flattening of the scalp EEG power spectrum also appears in depth SEEG recordings from adults with epilepsy. We used archived channel features from OpenNeuro ds004100 (HUP; 32 patients) and ds003876 version 1.0.2 (NIH, JHU, MIAMI, UMF; 29 patients), plus participant ages. Features came from the first 120 s of one interictal clip per patient, using within-shaft bipolar pairs; the primary patient value was the median across non-seizure-onset-zone pairs. In 60 patients aged 16–62 years, the site-centred Spearman association between age and the 30–45 Hz log-log exponent was positive (rho = +0.24, permutation p = 0.071, 95% CI −0.04 to +0.50), opposite to the predicted flattening. No leave-one-site-out estimate was negative. The 2–40 Hz exponent, 1–40 Hz power, and 80–150 Hz power showed no age association. Relative delta and alpha power had uncorrected associations (rho = −0.29 and +0.29), but neither passed Holm correction (p = 0.16). This cross-sectional result is limited by the small sample, few participants over 55, drug-resistant epilepsy and medication in all participants, uncontrolled electrode locations, bipolar montage, unknown sleep state, a single 120-s clip, and a narrow-band simple slope rather than a full aperiodic model.

SEEG · epilepsy · aperiodic exponent · age · interictal · power spectrum · bipolar electrodes · electroencephalography

Admitted 2026-11-09 16:39 UTC4 citing publications

References

usesThe 30–45 Hz aperiodic exponent did not separate seizure-onset channels in 31 SEEG patients sci:sha256:970015c7ad657e77ac8f470146719405ea5536dba9b4dabf2cdf20f1f8c23615
We used its archived per-channel interictal features (results/channel_features.tsv: exp_30_45, exp_2_40_robust, band powers; 32 HUP SEEG patients, ds004100), SOZ labels and participants.tsv (age, age_onset, target) as input for the age analysis.
usesAcross 29 SEEG patients, the 30–45 Hz exponent did not separate SOZ channels, while four other features did sci:sha256:115bc8cb0b4c9e536b5f083843a2745508d3c6635a8156b661a9e084f51582a7
We used its archived per-channel interictal features for 29 ds003876 SEEG patients (same code as 5ec6e903), SOZ labels, and participants.tsv (age, site) as the second cohort.
usesdoi:10.18112/openneuro.ds004100.v1.1.3
The HUP cohort contributed archived channel features, participant ages, seizure-onset-zone labels, and sampling metadata for 32 SEEG patients.
usesdoi:10.18112/openneuro.ds003876.v1.0.2
The NIH, JHU, MIAMI, and UMF cohort contributed archived channel features, participant ages, and seizure-onset-zone labels for the second cohort.

Cited by

4 publications

Public catalog entries on Sciscape.

supportsLocation control does not reveal age-related flattening of the 30–45 Hz SEEG aperiodic exponent in adults with epilepsy
After control for implant location, the adult 30-45 Hz exponent still did not become flatter with age (255 RAM patients, rho +0.14); the sign is positive as in this 60-patient HUP result (rho +0.24).
supportsNo moderate age-related flattening was detected in NREM non-SOZ bipolar SEEG
It found that the 30-45 Hz exponent of non-SOZ bipolar SEEG pairs does not flatten with age in 60 adults (16-62 y, ds004100 and ds003876). We used the same feature in different datasets that include 30 patients younger than 18 y, and also found no flattening (rho +0.06, CI -0.22 to +0.31).
supportsIn 261 RAM epilepsy patients aged 18–64, the 30–45 Hz SEEG exponent did not flatten with age (discovery rho +0.08; replication +0.22)
A larger sample (261 vs 60) also shows no flattening of the 30-45 Hz exponent with age; the sign is positive in both.
usesA HUP-trained interictal SEEG classifier ranked SOZ channels across four centers but did not outperform spike rate
We copied its inputs/ tables (per-channel features and participants for ds004100 and ds003876, with SHA256SUMS) and used them as the only input data.