Multi-Centre, Multi-Vendor & Multi-Disease Cardiac Image Segmentation Challenge
A fundamental step in cardiac imaging research is the annotation of relevant regions of the heart by an expert observer, that allow the extraction of important biomarkers. As this process is very inefficient when done manually, artificial intelligence has been proposed as a promising technique for its automation. In recent years, many machine/deep learning models have been proposed to accurately segment cardiac structures in magnetic resonance imaging. However, when these models are tested on unseen datasets acquired from distinct MRI scanners or clinical centres, the segmentation accuracy can be greatly reduced, a problem called lack of generalisation ability (i.e., the decrease of performance when a model is evaluated on a new centre or for a new scanner manufacturer).
The M&Ms challenge has been proposed to motivate the computer science community to develop generalisable CMR segmentation models that can be applied consistently across clinical centres. M&Ms will provide a reference dataset for the community, based on a large collection of studies from 6 different centres with 4 different magnetic resonance manufacturers. This collective effort is thought to yield models that can be open-sourced to the scientific community through the euCanSHare platform with good reliability in order to shorten the time needed for cardiac studies. Amongst others, a number of euCanSHare partners have contributed to the challenge, including Universitat de Barcelona (Project Coordination), McGill University Health Centre and Universitätsklinikum Hamburg-Eppendorf.
Winners will be announced on 4 October 2020 in the context of the 23rd International Conference on Medical Image Computing and Computer-Assisted Intervention (MICCAI 2020), to be held from October 4th to 8th, 2020 in Lima, Peru. The best three performing methods will receive Amazon Vouchers of 500, 300 and 200 euros, respectively. MICCAI 2020 is organized in collaboration with Pontifical Catholic University of Peru (PUCP) and attracts world-leading biomedical scientists, engineers, and clinicians from a wide range of disciplines associated with medical imaging and computer-assisted intervention. The conference series includes three days of oral presentations and poster sessions. MICCAI 2020 will also include workshops, tutorials, and challenges on the days preceding and succeeding the conference. These satellite events will offer a comprehensive forum to further explore topics relevant to MICCAI.