AUTHOR=Banerjee Ayan , Kamboj Payal , Wyckoff Sarah N. , Sussman Bethany L. , Gupta Sandeep K. S. , Boerwinkle Varina L. 

TITLE=Automated seizure onset zone locator from resting-state functional MRI in drug-resistant epilepsy

JOURNAL=Frontiers in Neuroimaging

VOLUME=Volume 1 - 2022

YEAR=2023

URL=https://www.frontiersin.org/journals/neuroimaging/articles/10.3389/fnimg.2022.1007668

DOI=10.3389/fnimg.2022.1007668

ISSN=2813-1193

ABSTRACT=Objective: Accurate localization of seizure onset zone (SOZ) from independent components (IC) of resting state functional magnetic resonance imaging (rs-fMRI) improves surgical outcomes in children with Drug Resistant Epilepsy (DRE). Automated IC sorting has limited success in identifying SOZ localizing ICs in adult normal rs-fMRI or uncategorized epilepsy. Children provide unique challenges due to the developing brain and associated surgical risks. This paper proposes a novel SOZ localization algorithm (EPIK) for children with DRE.
Methods: EPIK is developed in a phased approach, where fMRI noise-related biomarkers are used through high fidelity image processing techniques to eliminate noise ICs. Then SOZ markers are used through a maximum likelihood-based classifier to determine SOZ localizing ICs. The performance of EPIK was evaluated on a unique pediatric DRE dataset (n = 52). 24 children underwent surgical resection or ablation of rs-fMRI identified SOZ, concurrently evaluated with EEG and anatomical MRI. Two state-of-art techniques are used for comparison: a) least squares support vector machine, and b) Convolutional Neural Networks. The performance was benchmarked against expert IC sorting and Engel outcomes for surgical SOZ resection or ablation. Analysis is stratified across age and sex.  
Results: EPIK outperforms state-of-art techniques for SOZ localizing IC identification with mean accuracy of 84.7% (4% higher), precision 74.1% (22% higher), specificity 81.9% (3.2% higher) and sensitivity 88.6% (16.5% higher). EPIK provides consistent performance across age and sex with best performance in < 5-year age. It achieves a ~5-fold reduction in the number of ICs to be potentially analyzed during pre-surgical screening.