DeepLesionBrain: Towards a broader deep-learning generalization for multiple sclerosis lesion segmentation.

Reda Abdellah Kamraoui, Vinh-Thong Ta, Thomas Tourdias, Boris Mansencal, José V Manjon, Pierrick Coupé
Medical Image Analysis. 2022-02-01; 76: 102312
DOI: 10.1016/

Read on PubMed

Kamraoui RA(1), Ta VT(2), Tourdias T(3), Mansencal B(2), Manjon JV(4), Coup P(2).

Author information:
(1)Univ. Bordeaux, Bordeaux INP, CNRS, LaBRI, UMR5800, PICTURA, F-33400 Talence,
France. Electronic address: .
(2)Univ. Bordeaux, Bordeaux INP, CNRS, LaBRI, UMR5800, PICTURA, F-33400 Talence, France.
(3)Service de Neuroimagerie Diagnostique et Thrapeutique, Univ. Bordeaux,
F-33000 Bordeaux, France; Univ. Bordeaux, INSERM, Neurocentre Magendie,U1215,
F-3300 Bordeaux, France.
(4)ITACA, Universitat Politcnica de Valncia, 46022 Valencia, Spain.

Recently, segmentation methods based on Convolutional Neural Networks (CNNs)
showed promising performance in automatic Multiple Sclerosis (MS) lesions
segmentation. These techniques have even outperformed human experts in
controlled evaluation conditions such as Longitudinal MS Lesion Segmentation
Challenge (ISBI Challenge). However, state-of-the-art approaches trained to
perform well on highly-controlled datasets fail to generalize on clinical data
from unseen datasets. Instead of proposing another improvement of the
segmentation accuracy, we propose a novel method robust to domain shift and
performing well on unseen datasets, called DeepLesionBrain (DLB). This
generalization property results from three main contributions. First, DLB is
based on a large group of compact 3D CNNs. This spatially distributed strategy
aims to produce a robust prediction despite the risk of generalization failure
of some individual networks. Second, we propose a hierarchical specialization
learning (HSL) by pre-training a generic network over the whole brain, before
using its weights as initialization to locally specialized networks. By this
end, DLB learns both generic features extracted at global image level and
specific features extracted at local image level. Finally, DLB includes a new
image quality data augmentation to reduce dependency to training data
specificity (e.g., acquisition protocol). DLB generalization was validated in
cross-dataset experiments on MSSEG’16, ISBI challenge, and in-house datasets.
During experiments, DLB showed higher segmentation accuracy, better segmentation
consistency and greater generalization performance compared to state-of-the-art
methods. Therefore, DLB offers a robust framework well-suited for clinical

Copyright © 2021. Published by Elsevier B.V.


Know more about