Benoit Caldairou1, Niels A Foit1, Carlotta Mutti1, Fatemeh Fadaie1, Ravnoor Gill1, Hyo Min Lee1, Theo Demerath1, Horst Urbach1, Andreas Schulze-Bonhage1, Andrea Bernasconi2, Neda Bernasconi2. 1. From the Neuroimaging of Epilepsy Laboratory (B.C., N.A.F., C.M., F.F., R.G., H.M.L., A.B., N.B.), McConnell Brain Imaging Center, Montreal Neurological Institute, McGill University, Quebec, Canada; and Departments of Neurosurgery (N.A.F.) and Neuroradiology (T.D., H.U.), Freiburg Medical Center, and Department of Neurology (A.S.-B.), University of Freiburg, Germany. 2. From the Neuroimaging of Epilepsy Laboratory (B.C., N.A.F., C.M., F.F., R.G., H.M.L., A.B., N.B.), McConnell Brain Imaging Center, Montreal Neurological Institute, McGill University, Quebec, Canada; and Departments of Neurosurgery (N.A.F.) and Neuroradiology (T.D., H.U.), Freiburg Medical Center, and Department of Neurology (A.S.-B.), University of Freiburg, Germany. neda@bic.mni.mcgill.ca.
Abstract
BACKGROUND AND OBJECTIVES: MRI fails to reveal hippocampal pathology in 30% to 50% of temporal lobe epilepsy (TLE) surgical candidates. To address this clinical challenge, we developed an automated MRI-based classifier that lateralizes the side of covert hippocampal pathology in TLE. METHODS: We trained a surface-based linear discriminant classifier that uses T1-weighted (morphology) and T2-weighted and fluid-attenuated inversion recovery (FLAIR)/T1 (intensity) features. The classifier was trained on 60 patients with TLE (mean age 35.6 years, 58% female) with histologically verified hippocampal sclerosis (HS). Images were deemed to be MRI negative in 42% of cases on the basis of neuroradiologic reading (40% based on hippocampal volumetry). The predictive model automatically labeled patients as having left or right TLE. Lateralization accuracy was compared to electroclinical data, including side of surgery. Accuracy of the classifier was further assessed in 2 independent TLE cohorts with similar demographics and electroclinical characteristics (n = 57, 58% MRI negative). RESULTS: The overall lateralization accuracy was 93% (95% confidence interval 92%-94%), regardless of HS visibility. In MRI-negative TLE, the combination of T2 and FLAIR/T1 intensities provided the highest accuracy in both the training (84%, area under the curve [AUC] 0.95 ± 0.02) and validation (cohort 1 90%, AUC 0.99; cohort 2 76%, AUC 0.94) cohorts. DISCUSSION: This prediction model for TLE lateralization operates on readily available conventional MRI contrasts and offers gain in accuracy over visual radiologic assessment. The combined contribution of decreased T1- and increased T2-weighted intensities makes the synthetic FLAIR/T1 contrast particularly effective in MRI-negative HS, setting the basis for broad clinical translation. CLASSIFICATION OF EVIDENCE: This study provides Class II evidence that in people with TLE and MRI-negative HS, an automated MRI-based classifier accurately determines the side of pathology.
BACKGROUND AND OBJECTIVES: MRI fails to reveal hippocampal pathology in 30% to 50% of temporal lobe epilepsy (TLE) surgical candidates. To address this clinical challenge, we developed an automated MRI-based classifier that lateralizes the side of covert hippocampal pathology in TLE. METHODS: We trained a surface-based linear discriminant classifier that uses T1-weighted (morphology) and T2-weighted and fluid-attenuated inversion recovery (FLAIR)/T1 (intensity) features. The classifier was trained on 60 patients with TLE (mean age 35.6 years, 58% female) with histologically verified hippocampal sclerosis (HS). Images were deemed to be MRI negative in 42% of cases on the basis of neuroradiologic reading (40% based on hippocampal volumetry). The predictive model automatically labeled patients as having left or right TLE. Lateralization accuracy was compared to electroclinical data, including side of surgery. Accuracy of the classifier was further assessed in 2 independent TLE cohorts with similar demographics and electroclinical characteristics (n = 57, 58% MRI negative). RESULTS: The overall lateralization accuracy was 93% (95% confidence interval 92%-94%), regardless of HS visibility. In MRI-negative TLE, the combination of T2 and FLAIR/T1 intensities provided the highest accuracy in both the training (84%, area under the curve [AUC] 0.95 ± 0.02) and validation (cohort 1 90%, AUC 0.99; cohort 2 76%, AUC 0.94) cohorts. DISCUSSION: This prediction model for TLE lateralization operates on readily available conventional MRI contrasts and offers gain in accuracy over visual radiologic assessment. The combined contribution of decreased T1- and increased T2-weighted intensities makes the synthetic FLAIR/T1 contrast particularly effective in MRI-negative HS, setting the basis for broad clinical translation. CLASSIFICATION OF EVIDENCE: This study provides Class II evidence that in people with TLE and MRI-negative HS, an automated MRI-based classifier accurately determines the side of pathology.
Authors: Andrea Bernasconi; Fernando Cendes; William H Theodore; Ravnoor S Gill; Matthias J Koepp; Robert Edward Hogan; Graeme D Jackson; Paolo Federico; Angelo Labate; Anna Elisabetta Vaudano; Ingmar Blümcke; Philippe Ryvlin; Neda Bernasconi Journal: Epilepsia Date: 2019-05-28 Impact factor: 5.864
Authors: Lorenzo Caciagli; Andrea Bernasconi; Samuel Wiebe; Matthias J Koepp; Neda Bernasconi; Boris C Bernhardt Journal: Neurology Date: 2017-07-07 Impact factor: 9.910
Authors: Niels K Focke; Silvia B Bonelli; Mahinda Yogarajah; Catherine Scott; Mark R Symms; John S Duncan Journal: Epilepsia Date: 2009-03-09 Impact factor: 5.864
Authors: Daniel S Barron; Peter T Fox; Heath Pardoe; Jack Lancaster; Larry R Price; Karen Blackmon; Kristen Berry; Jose E Cavazos; Ruben Kuzniecky; Orrin Devinsky; Thomas Thesen Journal: Neuroimage Clin Date: 2014-08-07 Impact factor: 4.881
Authors: Gavin P Winston; Sjoerd B Vos; Jane L Burdett; M Jorge Cardoso; Sebastien Ourselin; John S Duncan Journal: Epilepsia Date: 2017-07-12 Impact factor: 5.864
Authors: Jessie Kulaga-Yoskovitz; Boris C Bernhardt; Seok-Jun Hong; Tommaso Mansi; Kevin E Liang; Andre J W van der Kouwe; Jonathan Smallwood; Andrea Bernasconi; Neda Bernasconi Journal: Sci Data Date: 2015-11-10 Impact factor: 6.444
Authors: Peter N Hadar; Lohith G Kini; Carlos Coto; Virginie Piskin; Lauren E Callans; Stephanie H Chen; Joel M Stein; Sandhitsu R Das; Paul A Yushkevich; Kathryn A Davis Journal: Neuroimage Clin Date: 2018-10-10 Impact factor: 4.881