Automated detection of focal cortical dysplasia in epilepsy using machine learning on structural MRI

Focal cortical dysplasia (FCD) represents a structural brain anomaly of developmental origin that significantly contributes to epilepsy. This proof-of-concept study evaluates the feasibility of machine learning (ML) algorithms to distinguish patients with FCD type II-associated epilepsy from healthy controls (HCs) using high-resolution structural MRI. Structural MRI data from a publicly available dataset comprising 167 subjects (both FCD-associated epilepsy patients and HCs) were analyzed.

Neuroanatomical features were extracted with FreeSurfer software. Preprocessing steps included dataset reconstruction, feature scaling, dimensionality reduction, and encoding of demographic variables. Multiple supervised ML models—including Logistic Regression, Support Vector Machine, Gradient Boosting, and Random Forest—were trained and evaluated using nested cross-validation with 10 outer folds and 5 inner folds.

Among the evaluated classifiers, Random Forest achieved the highest AUC (0.791). This model demonstrated balanced classification ability, as reflected by its ROC and Precision-Recall curves. While promising, these results reflect the model’s performance on the current dataset and require further validation.

This study demonstrates the preliminary potential of ML techniques to distinguish FCD-associated epilepsy from HCs using structural MRI, providing supportive information for clinical review. However, further studies with larger, independent clinical cohorts are necessary to validate these results before considering any clinical implementation. The authors would like to express their sincere appreciation to the Interdisciplinary Schools Institute for its invaluable support and for providing a comprehensive research platform.

School of Medicine, Tehran University of Medical Sciences, Tehran, Iran Department of Cognitive Neuroscience, Faculty of Interdisciplinary Sciences and Technologies, Tarbiat Modares University, Tehran, Iran Department of Neuroscience, Faculty of Medicine, Mashhad University of Medical Sciences, Mashhad, Iran Department of Review Studies, Interdisciplinary Schools Institute, Tehran, Iran Finetech in Medicine Research Center, Medical Physics Department, School of Medicine, Iran University of Medical Sciences, Tehran, Iran Department of Radiology, School of Paramedical Sciences, Gerash University of Medical Sciences, PO Box 7441758666, Gerash, Iran The authors declare no competing interests. This study used a publicly available and fully anonymized dataset from the OpenNeuro repository (Dataset ID: ds004199; https://openneuro.org/datasets/ds004199). All data collection, ethical approvals, and participant consent were obtained by the original investigators before public release.

The Research Ethics Committee of Gerash University of Medical Sciences (IR.GERUMS.REC.1403.022) reviewed and approved the conduct of this secondary analysis. No additional informed consent was required, as only de-identified public data were used. Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.