Related Experiment Video
Updated: Mar 30, 2026
A Multimodal Imaging- and Stimulation-based Method of Evaluating Connectivity-related Brain Excitability in Patients with Epilepsy
Published on: November 13, 2016
Automated MRI-based atrophy patterns to improve the diagnosis of generalized epilepsies in outpatient care
Thomas Checkouri1, Lois Laidebeur2, Jeremy Deverdun2
1Epilepsy Monitoring Unit, Department of Neurosurgery, Gui de Chauliac University Hospital, Montpellier, France; Centre National de la Recherche Scientifique (CNRS), Institute of Functional Genomics, University of Montpellier, Institut National de la Santé et de la Recherche Médicale (INSERM), France; Department of Neuroradiology, Montpellier-Nimes University Hospital, France.
Background:
Even after complete anamnestic, electroencephalographic and brain imaging evaluation, diagnostic uncertainty may persist in differentiating generalized genetic epilepsies (GGE) from focal epilepsies (FE), leading to errors in diagnosis and treatment. Based on recent findings about network-based patterns of progressive atrophy in most forms of epilepsy, we sought to evaluate the value of an automated morphometric analysis of gray and white matter to improve epilepsy classification in clinical practice.
Methods:
This prospective monocentric study evaluated atrophy patterns on 3D T1-weighted MRI analysed using a commercially available software (NeuroDeg sequence). Atrophy patterns (bilateral and symmetrical thalamo-frontal for GGE versus asymmetrical lobar or thalamic for FE) were compared to clinical, EEG, and imaging variables using multivariate logistic regression.
Results:
We included 44 patients in this study, among whom 20 were diagnosed with GGE and 24 with FE, based on expert epileptologist consensus. A bilateral and symmetrical atrophy pattern involving the thalami and/or frontal white matter was independently associated with GGE (p=0.043), yielding an AUC of 0.911 when combined with EEG and clinical variables. In contrast, the "Focal atrophy" pattern added no significant diagnosis value.
Discussion:
This study supports the use of an automated morphometric analysis as a clinically accessible tool to assist the diagnosis and treatment of GGE, representing a first step in the translation of connectome-based biomarkers in outpatient clinical practice.
