Multiple sclerosis lesions segmentation from multiple experts: The MICCAI 2016 challenge dataset.
NeuroImage. 2021-12-01; 244: 118589
DOI: 10.1016/j.neuroimage.2021.118589
Lire sur PubMed
Commowick O(1), Kain M(2), Casey R(3), Ameli R(3), Ferré JC(4), Kerbrat A(5), Tourdias T(6), Cervenansky F(7), Camarasu-Pop S(7), Glatard T(8), Vukusic S(3), Edan G(9), Barillot C(2), Dojat M(10), Cotton F(3).
Author information:
(1)Univ Rennes, Inria, CNRS, Inserm – IRISA UMR 6074, Empenn ERL U1228, Rennes
F-35000, France. Electronic address: https://olivier.commowick.org.
(2)Univ Rennes, Inria, CNRS, Inserm – IRISA UMR 6074, Empenn ERL U1228, Rennes
F-35000, France.
(3)Department of Radiology, Lyon Sud Hospital, Hospices Civils de Lyon, Lyon,
France.
(4)Univ Rennes, Inria, CNRS, Inserm – IRISA UMR 6074, Empenn ERL U1228, Rennes
F-35000, France; Department of Neuroradiology, CHU Rennes, Rennes F-35033,
France.
(5)Department of Neurology, CHU Rennes, Rennes F-35033, France.
(6)CHU de Bordeaux, Service de Neuro-Imagerie, Bordeaux, France.
(7)Univ Lyon, INSA-Lyon, Université Claude Bernard Lyon 1, UJM-Saint Etienne,
CNRS, Inserm, CREATIS UMR 5220, Lyon U1206, F-69621, France.
(8)Department of Computer Science and Software Engineering, Concordia
University, Montreal, Canada.
(9)Univ Rennes, Inria, CNRS, Inserm – IRISA UMR 6074, Empenn ERL U1228, Rennes
F-35000, France; Department of Neurology, CHU Rennes, Rennes F-35033, France.
(10)Inserm U1216, University Grenoble Alpes, CHU Grenoble, GIN, Grenoble,
France.
MRI plays a crucial role in multiple sclerosis diagnostic and patient follow-up.
In particular, the delineation of T2-FLAIR hyperintense lesions is crucial
although mostly performed manually – a tedious task. Many methods have thus been
proposed to automate this task. However, sufficiently large datasets with a
thorough expert manual segmentation are still lacking to evaluate these methods.
We present a unique dataset for MS lesions segmentation evaluation. It consists
of 53 patients acquired on 4 different scanners with a harmonized protocol.
Hyperintense lesions on FLAIR were manually delineated on each patient by 7
experts with control on T2 sequence, and gathered in a consensus segmentation
for evaluation. We provide raw and preprocessed data and a split of the dataset
into training and testing data, the latter including data from a scanner not
present in the training dataset. We strongly believe that this dataset will
become a reference in MS lesions segmentation evaluation, allowing to evaluate
many aspects: evaluation of performance on unseen scanner, comparison to
individual experts performance, comparison to other challengers who already used
this dataset, etc.
Copyright © 2021. Published by Elsevier Inc.
DOI: 10.1016/j.neuroimage.2021.118589
PMID: 34563682 [Indexed for MEDLINE]