Conférence Imagerie Médicale et Deep Learning - 8 septembre 2017

01 / 01 / 1970, Arnaud RENARD

Actualité

Le Prof Mattias HEINRICH, de l'université de Lübeck (Allemagne), fera une présentation vendredi 8 septembre 2017 à 11h en salle de radiologie sur le sujet

Multi-Atlas Segmentation of Cardiac MRI using Discrete Registration and its Relations to Deep Learning

About Prof Heinrich

Mattias Heinrich is Assistant Professor for Medical Image Analysis at the Institute of Medical Informatics at the University of Lübeck. His research focuses on developing novel methodologies for image analysis including discrete deformable image registration with application to multi-modal fusion and atlas-based segmentation, which has led to world-leading accuracies for abdominal CT and MRI cardiac segmentation as evident from two recent MICCAI challenges. Furthermore, together with his small team of 4 PhD students, he is developing (deep) machine learning algorithms for metric learning, action recognition, tracking, classification and semantic segmentation in medicine. He has published 13 articles in high impact journals such as Medical Image Analysis, NeuroImage, and IEEE Transactions on Medical Imaging, and a further 13 papers at the premier international conference on medical image analysis (MICCAI). Upon finishing his DPhil in Oxford, he won the BMVA Sullivan Award 2014 for the best thesis in the field of computer vision in the UK. He has co-organised a special session on ‘Image registration’ at the IEEE International Symposium on Biomedical Imaging (ISBI) 2015 in Brooklyn, NYC with over 100 attendees and the Workshop on Biomedical Image Registration (WBIR) 2017 in conjunction with CVPR in Las Vegas. He serves as reviewer or part of the program committees of nearly all vision and medical imaging conference and most major journals in the field. Mattias is a great believer in open-access and publishes his code at mpheinrich.de and github.com/mattiaspaul

http://www.mpheinrich.de/