Ravnoor Singh Gill1, Hyo-Min Lee1, Benoit Caldairou1, Seok-Jun Hong1, Carmen Barba1, Francesco Deleo1, Ludovico D'Incerti1, Vanessa Cristina Mendes Coelho1, Matteo Lenge1, Mira Semmelroch1, Dewi Victoria Schrader1, Fabrice Bartolomei1, Maxime Guye1, Andreas Schulze-Bonhage1, Horst Urbach1, Kyoo Ho Cho1, Fernando Cendes1, Renzo Guerrini1, Graeme Jackson1, R Edward Hogan1, Neda Bernasconi1, Andrea Bernasconi2. 1. From the Neuroimaging of Epilepsy Laboratory (R.S.G., H.-M.L., B.C., S.-J.H., N.B., A.B.), Montreal Neurological Institute, McGill University, Montreal, Quebec, Canada; Pediatric Neurology Unit and Laboratories (C.B., M.L., R.G.), Children's Hospital A. Meyer-University of Florence, Italy; Epilepsy Unit (F.D.) and Neuroradiology (L.D.), Fondazione IRCCS Istituto Neurologico C. Besta, Milan, Italy; Department of Neurology (V.C.M.C., F.C.), University of Campinas, Brazil; The Florey Institute of Neuroscience and Mental Health and The University of Melbourne (M.S., G.J.), Victoria, Australia; Department of Pediatrics (D.V.S.), British Columbia Children's Hospital, Vancouver, Canada; Aix Marseille University (F.B.), INSERM UMR 1106, Institut de Neurosciences des Systèmes; Aix Marseille University (M.G.), CNRS, CRMBM UMR 7339, Marseille, France; Freiburg Epilepsy Center (A.S.-B., H.U.), Universitätsklinikum Freiburg, Germany; Department of Neurology (K.H.C.), Yonsei University College of Medicine, Seoul, Korea; and Department of Neurology (R.E.H.), Washington University School of Medicine, St. Louis, MO. 2. From the Neuroimaging of Epilepsy Laboratory (R.S.G., H.-M.L., B.C., S.-J.H., N.B., A.B.), Montreal Neurological Institute, McGill University, Montreal, Quebec, Canada; Pediatric Neurology Unit and Laboratories (C.B., M.L., R.G.), Children's Hospital A. Meyer-University of Florence, Italy; Epilepsy Unit (F.D.) and Neuroradiology (L.D.), Fondazione IRCCS Istituto Neurologico C. Besta, Milan, Italy; Department of Neurology (V.C.M.C., F.C.), University of Campinas, Brazil; The Florey Institute of Neuroscience and Mental Health and The University of Melbourne (M.S., G.J.), Victoria, Australia; Department of Pediatrics (D.V.S.), British Columbia Children's Hospital, Vancouver, Canada; Aix Marseille University (F.B.), INSERM UMR 1106, Institut de Neurosciences des Systèmes; Aix Marseille University (M.G.), CNRS, CRMBM UMR 7339, Marseille, France; Freiburg Epilepsy Center (A.S.-B., H.U.), Universitätsklinikum Freiburg, Germany; Department of Neurology (K.H.C.), Yonsei University College of Medicine, Seoul, Korea; and Department of Neurology (R.E.H.), Washington University School of Medicine, St. Louis, MO. andrea.bernasconi@mcgill.ca.
Abstract
BACKGROUND AND OBJECTIVE: To test the hypothesis that a multicenter-validated computer deep learning algorithm detects MRI-negative focal cortical dysplasia (FCD). METHODS: We used clinically acquired 3-dimensional (3D) T1-weighted and 3D fluid-attenuated inversion recovery MRI of 148 patients (median age 23 years [range 2-55 years]; 47% female) with histologically verified FCD at 9 centers to train a deep convolutional neural network (CNN) classifier. Images were initially deemed MRI-negative in 51% of patients, in whom intracranial EEG determined the focus. For risk stratification, the CNN incorporated bayesian uncertainty estimation as a measure of confidence. To evaluate performance, detection maps were compared to expert FCD manual labels. Sensitivity was tested in an independent cohort of 23 cases with FCD (13 ± 10 years). Applying the algorithm to 42 healthy controls and 89 controls with temporal lobe epilepsy disease tested specificity. RESULTS: Overall sensitivity was 93% (137 of 148 FCD detected) using a leave-one-site-out cross-validation, with an average of 6 false positives per patient. Sensitivity in MRI-negative FCD was 85%. In 73% of patients, the FCD was among the clusters with the highest confidence; in half, it ranked the highest. Sensitivity in the independent cohort was 83% (19 of 23; average of 5 false positives per patient). Specificity was 89% in healthy and disease controls. DISCUSSION: This first multicenter-validated deep learning detection algorithm yields the highest sensitivity to date in MRI-negative FCD. By pairing predictions with risk stratification, this classifier may assist clinicians in adjusting hypotheses relative to other tests, increasing diagnostic confidence. Moreover, generalizability across age and MRI hardware makes this approach ideal for presurgical evaluation of MRI-negative epilepsy. CLASSIFICATION OF EVIDENCE: This study provides Class III evidence that deep learning on multimodal MRI accurately identifies FCD in patients with epilepsy initially diagnosed as MRI negative.
BACKGROUND AND OBJECTIVE: To test the hypothesis that a multicenter-validated computer deep learning algorithm detects MRI-negative focal cortical dysplasia (FCD). METHODS: We used clinically acquired 3-dimensional (3D) T1-weighted and 3D fluid-attenuated inversion recovery MRI of 148 patients (median age 23 years [range 2-55 years]; 47% female) with histologically verified FCD at 9 centers to train a deep convolutional neural network (CNN) classifier. Images were initially deemed MRI-negative in 51% of patients, in whom intracranial EEG determined the focus. For risk stratification, the CNN incorporated bayesian uncertainty estimation as a measure of confidence. To evaluate performance, detection maps were compared to expert FCD manual labels. Sensitivity was tested in an independent cohort of 23 cases with FCD (13 ± 10 years). Applying the algorithm to 42 healthy controls and 89 controls with temporal lobe epilepsy disease tested specificity. RESULTS: Overall sensitivity was 93% (137 of 148 FCD detected) using a leave-one-site-out cross-validation, with an average of 6 false positives per patient. Sensitivity in MRI-negative FCD was 85%. In 73% of patients, the FCD was among the clusters with the highest confidence; in half, it ranked the highest. Sensitivity in the independent cohort was 83% (19 of 23; average of 5 false positives per patient). Specificity was 89% in healthy and disease controls. DISCUSSION: This first multicenter-validated deep learning detection algorithm yields the highest sensitivity to date in MRI-negative FCD. By pairing predictions with risk stratification, this classifier may assist clinicians in adjusting hypotheses relative to other tests, increasing diagnostic confidence. Moreover, generalizability across age and MRI hardware makes this approach ideal for presurgical evaluation of MRI-negative epilepsy. CLASSIFICATION OF EVIDENCE: This study provides Class III evidence that deep learning on multimodal MRI accurately identifies FCD in patients with epilepsy initially diagnosed as MRI negative.
Authors: Geert Litjens; Thijs Kooi; Babak Ehteshami Bejnordi; Arnaud Arindra Adiyoso Setio; Francesco Ciompi; Mohsen Ghafoorian; Jeroen A W M van der Laak; Bram van Ginneken; Clara I Sánchez Journal: Med Image Anal Date: 2017-07-26 Impact factor: 8.545
Authors: Bilal Ahmed; Carla E Brodley; Karen E Blackmon; Ruben Kuzniecky; Gilad Barash; Chad Carlson; Brian T Quinn; Werner Doyle; Jacqueline French; Orrin Devinsky; Thomas Thesen Journal: Epilepsy Behav Date: 2015-05-31 Impact factor: 2.937
Authors: Renzo Guerrini; Michael Duchowny; Prasanna Jayakar; Pavel Krsek; Philippe Kahane; Laura Tassi; Federico Melani; Tilman Polster; Véronique M Andre; Carlos Cepeda; Darcy A Krueger; J Helen Cross; Roberto Spreafico; Mirco Cosottini; Jean Gotman; Francine Chassoux; Philippe Ryvlin; Fabrice Bartolomei; Andrea Bernasconi; Hermann Stefan; Ian Miller; Bertrand Devaux; Imad Najm; Flavio Giordano; Kristl Vonck; Carmen Barba; Ingmar Blumcke Journal: Epilepsia Date: 2015-10-05 Impact factor: 5.864
Authors: Susanne Fauser; Sanjay M Sisodiya; Lillian Martinian; Maria Thom; Christoph Gumbinger; Hans-Jürgen Huppertz; Claudia Hader; Karl Strobl; Bernhard J Steinhoff; Marco Prinz; Josef Zentner; Andreas Schulze-Bonhage Journal: Brain Date: 2009-06-08 Impact factor: 13.501
Authors: Thomas Thesen; Brian T Quinn; Chad Carlson; Orrin Devinsky; Jonathan DuBois; Carrie R McDonald; Jacqueline French; Richard Leventer; Olga Felsovalyi; Xiuyuan Wang; Eric Halgren; Ruben Kuzniecky Journal: PLoS One Date: 2011-02-04 Impact factor: 3.240
Authors: Sophie Adler; Konrad Wagstyl; Roxana Gunny; Lisa Ronan; David Carmichael; J Helen Cross; Paul C Fletcher; Torsten Baldeweg Journal: Neuroimage Clin Date: 2016-12-30 Impact factor: 4.881