Objective Evaluation of Multiple Sclerosis Lesion Segmentation using a Data Management and Processing Infrastructure

Commowick, Olivier
Istace, Audrey
Kain, Michaël
Laurent, Baptiste
Leray, Florent
Simon, Mathieu
Camarasu Pop, Sorina
Girard, Pascal
Améli, Roxana
Ferré, Jean-Christophe
Kerbrat, Anne
Tourdias, Thomas
Cervenansky, Frédéric
Glatard, Tristan
Beaumont, Jérémy
Doyle, Senan
Forbes, Florence
Knight, Jesse
Khademi, April
Mahbod, Amirreza
Wang, Chunliang
McKinley, Richard
Wagner, Franca
Muschelli, John
Sweeney, Elizabeth
Santos, Michel M.
Santos, Wellington P.
Silva-Filho, Abel G.
Tomás-Fernández, Xavier
Urien, Hélène
Bloch, Isabelle
Malpica, Norberto
Guttmann, Charles
Vukusic, Sandra
Edan, Gilles
Dojat, Michel
Styner, Martin
Warfield, Simon K.
Cotton, François
Barillot, Christian
We present a study of multiple sclerosis segmentation algorithms conducted at the international MICCAI 2016 challenge. This challenge was operated using a new open-science computing infrastructure. This allowed for the automatic and independent evaluation of a large range of algorithms in a fair and completely automatic manner. This computing infrastructure was used to evaluate thirteen methods of MS lesions segmentation, exploring a broad range of state-of-theart algorithms, against a high-quality database of 53 MS cases coming from four centers following a common definition of the acquisition protocol. Each case was annotated manually by an unprecedented number of seven different experts. Results of the challenge highlighted that automatic algorithms, including the recent machine learning methods (random forests, deep learning, …), are still trailing human expertise on both detection and delineation criteria. In addition, we demonstrate that computing a statistically robust consensus of the algorithms performs closer to human expertise on one score (segmentation) although still trailing on detection scores ​
This document is licensed under a Creative Commons:Attribution (by) Creative Commons by4.0