Symptom network analysis of prefrontal seizures.

Objective: Prefrontal seizures pose significant challenges in accurately identifying the complex interactions between clinical manifestations and brain electrophysiological activities. This proof-of-concept study aims to propose a new approach to rigorously support electroclinical reasoning in the field of epilepsy.

Methods: We analyzed stereoelectroencephalographic data from 42 patients with drug-resistant focal epilepsy, whose seizures involved prefrontal cortex at seizure onset. Semiological and brain activities features were scored by expert observers. We performed a symptom network analysis of semiological feature and a hybrid network analysis, coupling semiological features with network analysis of ictal brain activities. Centrality measures were used to identify the most influential features in the networks.

Results: Our analysis identified impairment of consciousness as the most central feature in the semiological network. In the hybrid network, the anterior cingulate area (here incorporating Brodmann area [BA]-32 and/or rostral part of BA-24) emerged as the most central brain activity feature.

Significance: By integrating semiological features with brain electrophysiological activities into hybrid networks, symptom network analysis offers an effective quantitative tool for examining the relationships between seizure semiology and brain activity correlates in prefrontal seizures. This study provides an opportunity to advance a novel approach to rigorously investigate the intricacies of electroclinical correlations, sustaining the development of dynamic models, on different series of focal epilepsies, larger cohorts, and semiological features automatically extracted by artificial intelligence, that better reflect the temporal and spatial complexities of seizure propagation in patients with complex seizures.