Individualised Functional Brain Mapping Distinguishes Drug-Resistant from Early-Stage Epilepsy

Epilepsy is among the most prevalent neurological disorders, affecting millions of individuals worldwide at every stage of life. Characterised by recurrent seizures, epilepsy can significantly disrupt daily functioning, education, employment, and overall quality of life. Despite advances in neuroimaging, current approaches often overlook the individualised nature of brain disruptions in epilepsy. Here, we introduce an individualised functional Magnetic Resonance Imaging (fMRI) framework, Adjusted Local Estimates of Connectivity (ALEC), to detect patient-specific brain local connectivity abnormalities across distinct clinical stages of epilepsy. To do so, we analysed movie-watching multi-echo fMRI in 102 heterogeneous epilepsy patients (34 with drug-resistant epilepsy; 34 with a new diagnosis of epilepsy; 34 after a first seizure) and 68 socioeconomically matched healthy controls. ALEC is a voxel-wise modified z-score and estimates deviations in Regional Homogeneity from healthy norms at the individual level. Our results show that whole-brain averaged ALEC scores were significantly higher in drug-resistant epilepsy compared to early-stage cohorts. Several drug-resistant individuals exhibited pronounced ALEC elevations in the hippocampus, thalamus and brainstem alongside widespread cortical decreases, although these patterns did not reach group-level significance. Age and seizure duration correlated positively with ALEC, but only within the drug-resistant group. We also highlight a subset of cases that demonstrated concordance with ALEC and the patients clinical history and investigations, including epileptogenic pathology. Combined, our findings highlight the importance of individualised neuroimaging approaches for understanding epilepsy. By revealing biologically concordant local connectivity patterns--marked by local hyperconnectivity in drug-resistant cases worsening with aging--ALEC provides a potential pathway for precision brain mapping of patients at risk for drug resistance.