Deep learning methods for automatic evaluation of delayed enhancement-MRI. The results of the EMIDEC challenge

Alain Lalande, Zhihao Chen, Thibaut Pommier, Thomas Decourselle, Abdul Qayyum, Michel Salomon, Dominique Ginhac, Youssef Skandarani, Arnaud Boucher, Khawla Brahim, Marleen de Bruijne, Robin Camarasa, Teresa M. Correia, Xue Feng, Kibrom Berihu Girum, Anja Hennemuth, Markus Hüllebrand, Raabid Hussain, Matthias Ivantsits, Jun Ma et al. (13 additional authors not shown)

Abstract: Highlights•Dataset of 150 DE-MRI exams in short-axis orientation with the manual drawing.•The used dataset include clinical information that could be recorded in emergency department in addition to the MR images.•The first objective is to compare the latest methodological developments in image processing to segment the DE-MRI exams.•The second objective is to automatically classify the exams into non-pathological and pathological (myocardial infarction).