Please use this identifier to cite or link to this item:
Title: Machine Learning Approaches for Imaging-Based Prognostication of the Outcome of Surgery for Mesial Temporal Lobe Epilepsy.
Austin Authors: Sinclair, Benjamin;Cahill, Varduhi;Seah, Jarrel;Kitchen, Andy;Vivash, Lucy E;Chen, Zhibin;Malpas, Charles B;O'Shea, Marie F ;Desmond, Patricia M;Hicks, Rodney J;Morokoff, Andrew P;King, James A;Fabinyi, Gavin C ;Kaye, Andrew H;Kwan, Patrick;Berkovic, Samuel F ;Law, Meng;O'Brien, Terence J
Affiliation: Department of Medicine, University of Melbourne, Melbourne, Victoria, Australia..
Epilepsy Research Centre
Surgery (University of Melbourne)
Department of Surgery, University of Melbourne, Royal Melbourne Hospital, Melbourne, Victoria, Australia..
Peter MacCallum Cancer Centre and the Sir Peter MacCallum Department of Oncology, University of Melbourne, Melbourne, Victoria, Australia..
Department of Radiology, University of Melbourne, Royal Melbourne Hospital, Melbourne, Victoria, Australia..
Comprehensive Epilepsy Program
Melbourne School of Psychological Sciences, University of Melbourne, Melbourne, Victoria, Australia..
Department of Radiology, Alfred Health, Melbourne, Victoria, Australia..
Department of Neurology, Melbourne Brain Centre, Royal Melbourne Hospital, Melbourne, Victoria, Australia..
Department Neurology, Alfred Health, Melbourne, Victoria, Australia..
Department of Neuroscience, Central Clinical School, Monash University, Melbourne, Victoria, Australia..
Academic Neurology Unit, University of Sheffield, Royal Hallamshire Hospital, Sheffield, United Kingdom..
Department of Neurosurgery, Hadassah Hebrew University Hospital, Jerusalem, Israel..
Division of Neuroscience and Experimental Psychology, School of Biological Sciences, University of Manchester, Manchester, United Kingdom..
Issue Date: 2022
Date: 2022-03-25
Publication information: Epilepsia 2022; 63(5): 1081-1092
Abstract: Around 30% of patients undergoing surgical resection for drug-resistant mesial temporal lobe epilepsy (MTLE) do not obtain seizure freedom. Success of anterior temporal lobe resection (ATLR) critically depends on the careful selection of surgical candidates, aiming at optimizing seizure freedom while minimizing postoperative morbidity. Structural MRI and FDG-PET neuroimaging are routinely used in presurgical assessment and guide the decision to proceed to surgery. In this study, we evaluate the potential of machine learning techniques (logistic regression, support vector machines, random forests and artificial neural networks) applied to standard presurgical MRI and PET imaging features to provide enhanced prognostic value relative to current practice. In the study cohort, 24/82 (28.3%) who underwent an ATLR for drug resistant MTLE did not achieve an Engel Class I (i.e. free of disabling seizures) outcome at a minimum of 2 years post-operative follow-up. We found that machine learning approaches were able to predict up to 73% of the 24 ATLR surgical patients who did not achieve a Class I outcome, at the expense of incorrect prediction for up to 31% of patients who did achieve a Class I outcome. Overall accuracies ranged from 70-80% and area under curve (AUC) of receiver operating characteristic of 0.75-0.81. We additionally found that information regarding overall extent of both total and significantly hypometabolic tissue resected was crucial to predictive performance, with AUC dropping to 0.59-0.62 using presurgical information alone. Incorporating the laterality of seizure onset and the choice of machine learning algorithm did not significantly change predictive performance. Collectively, these results indicate that 'acceptable' to 'good' patient specific prognostication for drug resistant MTLE surgery is feasible with machine learning approaches utilizing commonly collected imaging modalities, but that information on the surgical resection region is critical for optimal prognostication.
DOI: 10.1111/epi.17217
Journal: Epilepsia
PubMed URL: 35266138
PubMed URL:
Type: Journal Article
Subjects: Epilepsy
Machine Learning
Appears in Collections:Journal articles

Show full item record

Page view(s)

checked on May 30, 2024

Google ScholarTM


Items in AHRO are protected by copyright, with all rights reserved, unless otherwise indicated.