Cargando…

Online hard example mining vs. fixed oversampling strategy for segmentation of new multiple sclerosis lesions from longitudinal FLAIR MRI

Detecting new lesions is a key aspect of the radiological follow-up of patients with Multiple Sclerosis (MS), leading to eventual changes in their therapeutics. This paper presents our contribution to the MSSEG-2 MICCAI 2021 challenge. The challenge is focused on the segmentation of new MS lesions u...

Descripción completa

Detalles Bibliográficos
Autores principales: Schmidt-Mengin, Marius, Soulier, Théodore, Hamzaoui, Mariem, Yazdan-Panah, Arya, Bodini, Benedetta, Ayache, Nicholas, Stankoff, Bruno, Colliot, Olivier
Formato: Online Artículo Texto
Lenguaje:English
Publicado: Frontiers Media S.A. 2022
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9672803/
https://www.ncbi.nlm.nih.gov/pubmed/36408404
http://dx.doi.org/10.3389/fnins.2022.1004050
_version_ 1784832821224800256
author Schmidt-Mengin, Marius
Soulier, Théodore
Hamzaoui, Mariem
Yazdan-Panah, Arya
Bodini, Benedetta
Ayache, Nicholas
Stankoff, Bruno
Colliot, Olivier
author_facet Schmidt-Mengin, Marius
Soulier, Théodore
Hamzaoui, Mariem
Yazdan-Panah, Arya
Bodini, Benedetta
Ayache, Nicholas
Stankoff, Bruno
Colliot, Olivier
author_sort Schmidt-Mengin, Marius
collection PubMed
description Detecting new lesions is a key aspect of the radiological follow-up of patients with Multiple Sclerosis (MS), leading to eventual changes in their therapeutics. This paper presents our contribution to the MSSEG-2 MICCAI 2021 challenge. The challenge is focused on the segmentation of new MS lesions using two consecutive Fluid Attenuated Inversion Recovery (FLAIR) Magnetic Resonance Imaging (MRI). In other words, considering longitudinal data composed of two time points as input, the aim is to segment the lesional areas, which are present only in the follow-up scan and not in the baseline. The backbone of our segmentation method is a 3D UNet applied patch-wise to the images, and in which, to take into account both time points, we simply concatenate the baseline and follow-up images along the channel axis before passing them to the 3D UNet. Our key methodological contribution is the use of online hard example mining to address the challenge of class imbalance. Indeed, there are very few voxels belonging to new lesions which makes training deep-learning models difficult. Instead of using handcrafted priors like brain masks or multi-stage methods, we experiment with a novel modification to online hard example mining (OHEM), where we use an exponential moving average (i.e., its weights are updated with momentum) of the 3D UNet to mine hard examples. Using a moving average instead of the raw model should allow smoothing of its predictions and allow it to give more consistent feedback for OHEM.
format Online
Article
Text
id pubmed-9672803
institution National Center for Biotechnology Information
language English
publishDate 2022
publisher Frontiers Media S.A.
record_format MEDLINE/PubMed
spelling pubmed-96728032022-11-19 Online hard example mining vs. fixed oversampling strategy for segmentation of new multiple sclerosis lesions from longitudinal FLAIR MRI Schmidt-Mengin, Marius Soulier, Théodore Hamzaoui, Mariem Yazdan-Panah, Arya Bodini, Benedetta Ayache, Nicholas Stankoff, Bruno Colliot, Olivier Front Neurosci Neuroscience Detecting new lesions is a key aspect of the radiological follow-up of patients with Multiple Sclerosis (MS), leading to eventual changes in their therapeutics. This paper presents our contribution to the MSSEG-2 MICCAI 2021 challenge. The challenge is focused on the segmentation of new MS lesions using two consecutive Fluid Attenuated Inversion Recovery (FLAIR) Magnetic Resonance Imaging (MRI). In other words, considering longitudinal data composed of two time points as input, the aim is to segment the lesional areas, which are present only in the follow-up scan and not in the baseline. The backbone of our segmentation method is a 3D UNet applied patch-wise to the images, and in which, to take into account both time points, we simply concatenate the baseline and follow-up images along the channel axis before passing them to the 3D UNet. Our key methodological contribution is the use of online hard example mining to address the challenge of class imbalance. Indeed, there are very few voxels belonging to new lesions which makes training deep-learning models difficult. Instead of using handcrafted priors like brain masks or multi-stage methods, we experiment with a novel modification to online hard example mining (OHEM), where we use an exponential moving average (i.e., its weights are updated with momentum) of the 3D UNet to mine hard examples. Using a moving average instead of the raw model should allow smoothing of its predictions and allow it to give more consistent feedback for OHEM. Frontiers Media S.A. 2022-11-04 /pmc/articles/PMC9672803/ /pubmed/36408404 http://dx.doi.org/10.3389/fnins.2022.1004050 Text en Copyright © 2022 Schmidt-Mengin, Soulier, Hamzaoui, Yazdan-Panah, Bodini, Ayache, Stankoff and Colliot. https://creativecommons.org/licenses/by/4.0/This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
spellingShingle Neuroscience
Schmidt-Mengin, Marius
Soulier, Théodore
Hamzaoui, Mariem
Yazdan-Panah, Arya
Bodini, Benedetta
Ayache, Nicholas
Stankoff, Bruno
Colliot, Olivier
Online hard example mining vs. fixed oversampling strategy for segmentation of new multiple sclerosis lesions from longitudinal FLAIR MRI
title Online hard example mining vs. fixed oversampling strategy for segmentation of new multiple sclerosis lesions from longitudinal FLAIR MRI
title_full Online hard example mining vs. fixed oversampling strategy for segmentation of new multiple sclerosis lesions from longitudinal FLAIR MRI
title_fullStr Online hard example mining vs. fixed oversampling strategy for segmentation of new multiple sclerosis lesions from longitudinal FLAIR MRI
title_full_unstemmed Online hard example mining vs. fixed oversampling strategy for segmentation of new multiple sclerosis lesions from longitudinal FLAIR MRI
title_short Online hard example mining vs. fixed oversampling strategy for segmentation of new multiple sclerosis lesions from longitudinal FLAIR MRI
title_sort online hard example mining vs. fixed oversampling strategy for segmentation of new multiple sclerosis lesions from longitudinal flair mri
topic Neuroscience
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9672803/
https://www.ncbi.nlm.nih.gov/pubmed/36408404
http://dx.doi.org/10.3389/fnins.2022.1004050
work_keys_str_mv AT schmidtmenginmarius onlinehardexampleminingvsfixedoversamplingstrategyforsegmentationofnewmultiplesclerosislesionsfromlongitudinalflairmri
AT souliertheodore onlinehardexampleminingvsfixedoversamplingstrategyforsegmentationofnewmultiplesclerosislesionsfromlongitudinalflairmri
AT hamzaouimariem onlinehardexampleminingvsfixedoversamplingstrategyforsegmentationofnewmultiplesclerosislesionsfromlongitudinalflairmri
AT yazdanpanaharya onlinehardexampleminingvsfixedoversamplingstrategyforsegmentationofnewmultiplesclerosislesionsfromlongitudinalflairmri
AT bodinibenedetta onlinehardexampleminingvsfixedoversamplingstrategyforsegmentationofnewmultiplesclerosislesionsfromlongitudinalflairmri
AT ayachenicholas onlinehardexampleminingvsfixedoversamplingstrategyforsegmentationofnewmultiplesclerosislesionsfromlongitudinalflairmri
AT stankoffbruno onlinehardexampleminingvsfixedoversamplingstrategyforsegmentationofnewmultiplesclerosislesionsfromlongitudinalflairmri
AT colliotolivier onlinehardexampleminingvsfixedoversamplingstrategyforsegmentationofnewmultiplesclerosislesionsfromlongitudinalflairmri