Interpretable surface-based detection of focal cortical dysplasias: a Multi-centre Epilepsy Lesion Detection study

Spitzer H, Ripart M, Whitaker K, D'Arco F, Mankad K, Chen AA, Napolitano A, De Palma L, De Benedictis A, Foldes S, Humphreys Z, Zhang K, Hu W, Mo J, Likeman M, Davies S, Guttler C, Lenge M, Cohen NT, Tang Y, Wang S, Chari A, Tisdall M, Bargallo N, Conde-Blanco E, Pariente JC, Pascual-Diaz S, Delgado-Martinez I, Perez-Enriquez C, Lagorio I, Abela E, Mullatti N, O'Muircheartaigh J, Vecchiato K, Liu Y, Caligiuri ME, Sinclair B, Vivash L, Willard A, Kandasamy J, Mclellan A, Sokol D, Semmelroch M, Kloster AG, Opheim G, Ribeiro L, Yasuda C, Rossi-Espagnet C, Hamandi K, Tietze A, Barba C, Guerrini R, Gaillard WD, You X, Wang I, Gonzalez-Ortiz S, Severino M, Striano P, Tortora D, Kalviainen R, Gambardella A, Labate A, Desmond P, Lui E, O'Brien T, Shetty J, Jackson G, Duncan JS, Winston GP, Pinborg LH, Cendes F, Theis FJ, Shinohara RT, Cross JH, Baldeweg T, Adler S, Wagstyl K (2022)

Publication Type: Journal article

Publication year: 2022

Journal

Book Volume: 145

Pages Range: 3859-3871

Journal Issue: 11

DOI: 10.1093/brain/awac224

Abstract

One outstanding challenge for machine learning in diagnostic biomedical imaging is algorithm interpretability. A key application is the identification of subtle epileptogenic focal cortical dysplasias (FCDs) from structural MRI. FCDs are difficult to visualize on structural MRI but are often amenable to surgical resection. We aimed to develop an open-source, interpretable, surface-based machine-learning algorithm to automatically identify FCDs on heterogeneous structural MRI data from epilepsy surgery centres worldwide. The Multi-centre Epilepsy Lesion Detection (MELD) Project collated and harmonized a retrospective MRI cohort of 1015 participants, 618 patients with focal FCD-related epilepsy and 397 controls, from 22 epilepsy centres worldwide. We created a neural network for FCD detection based on 33 surface-based features. The network was trained and cross-validated on 50% of the total cohort and tested on the remaining 50% as well as on 2 independent test sites. Multidimensional feature analysis and integrated gradient saliencies were used to interrogate network performance. Our pipeline outputs individual patient reports, which identify the location of predicted lesions, alongside their imaging features and relative saliency to the classifier. On a restricted ‘gold-standard’ subcohort of seizure-free patients with FCD type IIB who had T1 and fluid-attenuated inversion recovery MRI data, the MELD FCD surface-based algorithm had a sensitivity of 85%. Across the entire withheld test cohort the sensitivity was 59% and specificity was 54%. After including a border zone around lesions, to account for uncertainty around the borders of manually delineated lesion masks, the sensitivity was 67%. This multicentre, multinational study with open access protocols and code has developed a robust and interpretable machine-learning algorithm for automated detection of focal cortical dysplasias, giving physicians greater confidence in the identification of subtle MRI lesions in individuals with epilepsy.

Involved external institutions

University of Eastern Finland Finland (FI) Magna Græcia University of Catanzaro / Università degli studi Magna Græcia di Catanzaro Italy (IT) Sichuan University (SCU) / 四川大学 China (CN) Universidade Estadual de Campinas (UNICAMP) / University of Campinas Brazil (BR) Ospedale Pediatrico Bambino Gesu Italy (IT) Cardiff University United Kingdom (GB) Charité - Universitätsmedizin Berlin Germany (DE) Università degli Studi di Firenze / University of Florence Italy (IT) Children’s National Health System United States (USA) (US) Cleveland Clinic United States (USA) (US) Hospital del Mar Spain (ES) Istituto Giannina Gaslini Italy (IT) Copenhagen University Hospital Denmark (DK) Helmholtz Zentrum München - Deutsches Forschungszentrum für Gesundheit und Umwelt (HMGU) / Helmholtz Munich Germany (DE) University of Pennsylvania United States (USA) (US) Great Ormond Street Hospital (GOSH) United Kingdom (GB) The Alan Turing Institute United Kingdom (GB) Phoenix Children's Hospital United States (USA) (US) Capital University of Medical Sciences / 首都医科大学 China (CN) Bristol Royal Hospital for Children (BRHC) United Kingdom (GB) Hospital Clínic de Barcelona Spain (ES) Institut d'investigacions Biomèdiques August Pi i Sunyer (IDIBAPS) Spain (ES) Psychiatrische Dienste Aargau AG Switzerland (CH) King’s College London United Kingdom (GB) Monash University Australia (AU) Royal Hospital for Children & Young People (RHCYP) United Kingdom (GB) The University of Melbourne Australia (AU) University of Messina / Università degli Studi di Messina Italy (IT) The Florey Institute of Neuroscience and Mental Health Australia (AU) University College London (UCL) United Kingdom (GB)

How to cite

APA:

Spitzer, H., Ripart, M., Whitaker, K., D'Arco, F., Mankad, K., Chen, A.A.,... Wagstyl, K. (2022). Interpretable surface-based detection of focal cortical dysplasias: a Multi-centre Epilepsy Lesion Detection study. Brain, 145(11), 3859-3871. https://doi.org/10.1093/brain/awac224

MLA:

Spitzer, Hannah, et al. "Interpretable surface-based detection of focal cortical dysplasias: a Multi-centre Epilepsy Lesion Detection study." Brain 145.11 (2022): 3859-3871.

BibTeX: Download