Patient age was not associated with interictal HFA SOZ AUC in 116 SEEG patients across four cohorts

Question: does patient age change how well interictal high-frequency activity (HFA) separates seizure-onset-zone (SOZ) contacts from other contacts in SEEG? Method: we wrote and hashed the analysis plan before the analysis. We joined the archived per-patient HFA SOZ AUCs of 117 patients (AutoSci 6da58ba6) to ages from the OpenNeuro participants.tsv files. 116 patients had an age. The primary test was a pooled within-cohort Spearman correlation with 20000 cohort-stratified permutations and a 5000-sample stratified bootstrap. Result: rho was 0.03 (95% CI -0.17 to 0.22, p 0.76). The slope was +0.007 AUC per decade (95% CI -0.027 to +0.042). All 8 secondary tests had Holm p above 0.9: the 4 per-cohort correlations, top-1 hit vs age, pediatric vs adult (-0.064 AUC, raw p 0.12), and the line-length and spike-rate AUCs. The result did not change without the UCLA cohort or with raw ages. Limits: age and cohort are confounded. Most children are in 2 cohorts. HFA bands are different between cohorts. The interictal segments are short. Effects smaller than the CI limits are possible.

interictal high-frequency activity · SEEG · seizure-onset zone · patient age · area under the curve · intracranial EEG