Utility of intracranial EEG networks depends on re-referencing and connectivity choice

• DOI: 10.1093/braincomms/fcae165
• Publication Date: 2024-05-13T07:41:26Z
• Authors: Haoer Shi; Akash Ranjan Patt...; Carlos Aguila; Alfredo Lucas; Nishant Sinha; Brian Prager; Marissa Mojena; Ryan Gallagher; Alexandra Parashos; Leonardo Bonilha; Ezequiel Gleichge...; Kathryn A Davis; Brian Litt; Erin C Conrad
• Keywords: 03 medical and health sciences; 0302 clinical medicine; Original Article

Studies of intracranial EEG networks have been used to reveal seizure generators in patients with drug-resistant epilepsy. Intracranial is implanted capture the epileptic network, collection brain tissue that forms a substrate for seizures start and spread. Interictal measures activity at baseline, computed during this state can aberrant without requiring recordings. network analyses require choosing reference applying statistical functional connectivity. Approaches these technical choices vary widely across studies, impact on downstream poorly understood. Our objective was examine effects different re-referencing connectivity approaches results ability lateralize onset zone We applied 48 pre-processing pipelines cohort 125 epilepsy recorded interictal two centres generate networks. Twenty-four time frequency domains were combination common average or bipolar re-referencing. an unsupervised clustering algorithm identify groups pipelines. subjected each approach three quality tests: (i) introduction spurious correlations; (ii) robustness incomplete spatial sampling; (iii) clinician-defined zone. Three similar emerged: pipelines, relative entropy-based Relative entropy more robust electrode sampling than other methods (Friedman test, Dunn-Šidák test