A semi-supervised large margin algorithm for white matter hyperintensity segmentation

Chen Qin, Ricardo Guerrero Moreno, Christopher Bowles, Christian Ledig, Philip Scheltens, Frederik Barkhof, Hanneke Rhodius-Meester, Betty Tijms, Afina W. Lemstra, Wiesje M. Van Der Flier, Ben Glocker, Daniel Rueckert

Research output: Contribution to conferencePaperOther research output

Abstract

Precise detection and quantification of white matter hyperintensities (WMH) is of great interest in studies of neurodegenerative diseases (NDs). In this work, we propose a novel semi-supervised large margin algorithm for the segmentation of WMH. The proposed algorithm optimizes a kernel based max-margin objective function which aims to maximize the margin averaged over inliers and outliers while exploiting a limited amount of available labelled data. We show that the learning problem can be formulated as a joint framework learning a classifier and a label assignment simultaneously, which can be solved efficiently by an iterative algorithm. We evaluate our method on a database of 280 brain Magnetic Resonance (MR) images from subjects that either suffered from subjective memory complaints or were diagnosed with NDs. The segmented WMH volumes correlate well with the standard clinical measurement (Fazekas score), and both the qualitative visualization results and quantitative correlation scores of the proposed algorithm outperform other well known methods for WMH segmentation.

Conference

Conference7th International Workshop on Machine Learning in Medical Imaging, MLMI 2016 held in conjunction with 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016
CountryGreece
CityAthens
Period17/10/201617/10/2016

Cite this

Qin, C., Moreno, R. G., Bowles, C., Ledig, C., Scheltens, P., Barkhof, F., ... Rueckert, D. (2016). A semi-supervised large margin algorithm for white matter hyperintensity segmentation. 104-112. Paper presented at 7th International Workshop on Machine Learning in Medical Imaging, MLMI 2016 held in conjunction with 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016, Athens, Greece. https://doi.org/10.1007/978-3-319-47157-0_13
Qin, Chen ; Moreno, Ricardo Guerrero ; Bowles, Christopher ; Ledig, Christian ; Scheltens, Philip ; Barkhof, Frederik ; Rhodius-Meester, Hanneke ; Tijms, Betty ; Lemstra, Afina W. ; Van Der Flier, Wiesje M. ; Glocker, Ben ; Rueckert, Daniel. / A semi-supervised large margin algorithm for white matter hyperintensity segmentation. Paper presented at 7th International Workshop on Machine Learning in Medical Imaging, MLMI 2016 held in conjunction with 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016, Athens, Greece.9 p.
@conference{ba64935e12a141c984223f805dd07b54,
title = "A semi-supervised large margin algorithm for white matter hyperintensity segmentation",
abstract = "Precise detection and quantification of white matter hyperintensities (WMH) is of great interest in studies of neurodegenerative diseases (NDs). In this work, we propose a novel semi-supervised large margin algorithm for the segmentation of WMH. The proposed algorithm optimizes a kernel based max-margin objective function which aims to maximize the margin averaged over inliers and outliers while exploiting a limited amount of available labelled data. We show that the learning problem can be formulated as a joint framework learning a classifier and a label assignment simultaneously, which can be solved efficiently by an iterative algorithm. We evaluate our method on a database of 280 brain Magnetic Resonance (MR) images from subjects that either suffered from subjective memory complaints or were diagnosed with NDs. The segmented WMH volumes correlate well with the standard clinical measurement (Fazekas score), and both the qualitative visualization results and quantitative correlation scores of the proposed algorithm outperform other well known methods for WMH segmentation.",
author = "Chen Qin and Moreno, {Ricardo Guerrero} and Christopher Bowles and Christian Ledig and Philip Scheltens and Frederik Barkhof and Hanneke Rhodius-Meester and Betty Tijms and Lemstra, {Afina W.} and {Van Der Flier}, {Wiesje M.} and Ben Glocker and Daniel Rueckert",
year = "2016",
doi = "10.1007/978-3-319-47157-0_13",
language = "English",
pages = "104--112",
note = "7th International Workshop on Machine Learning in Medical Imaging, MLMI 2016 held in conjunction with 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016 ; Conference date: 17-10-2016 Through 17-10-2016",

}

Qin, C, Moreno, RG, Bowles, C, Ledig, C, Scheltens, P, Barkhof, F, Rhodius-Meester, H, Tijms, B, Lemstra, AW, Van Der Flier, WM, Glocker, B & Rueckert, D 2016, 'A semi-supervised large margin algorithm for white matter hyperintensity segmentation' Paper presented at 7th International Workshop on Machine Learning in Medical Imaging, MLMI 2016 held in conjunction with 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016, Athens, Greece, 17/10/2016 - 17/10/2016, pp. 104-112. https://doi.org/10.1007/978-3-319-47157-0_13

A semi-supervised large margin algorithm for white matter hyperintensity segmentation. / Qin, Chen; Moreno, Ricardo Guerrero; Bowles, Christopher; Ledig, Christian; Scheltens, Philip; Barkhof, Frederik; Rhodius-Meester, Hanneke; Tijms, Betty; Lemstra, Afina W.; Van Der Flier, Wiesje M.; Glocker, Ben; Rueckert, Daniel.

2016. 104-112 Paper presented at 7th International Workshop on Machine Learning in Medical Imaging, MLMI 2016 held in conjunction with 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016, Athens, Greece.

Research output: Contribution to conferencePaperOther research output

TY - CONF

T1 - A semi-supervised large margin algorithm for white matter hyperintensity segmentation

AU - Qin, Chen

AU - Moreno, Ricardo Guerrero

AU - Bowles, Christopher

AU - Ledig, Christian

AU - Scheltens, Philip

AU - Barkhof, Frederik

AU - Rhodius-Meester, Hanneke

AU - Tijms, Betty

AU - Lemstra, Afina W.

AU - Van Der Flier, Wiesje M.

AU - Glocker, Ben

AU - Rueckert, Daniel

PY - 2016

Y1 - 2016

N2 - Precise detection and quantification of white matter hyperintensities (WMH) is of great interest in studies of neurodegenerative diseases (NDs). In this work, we propose a novel semi-supervised large margin algorithm for the segmentation of WMH. The proposed algorithm optimizes a kernel based max-margin objective function which aims to maximize the margin averaged over inliers and outliers while exploiting a limited amount of available labelled data. We show that the learning problem can be formulated as a joint framework learning a classifier and a label assignment simultaneously, which can be solved efficiently by an iterative algorithm. We evaluate our method on a database of 280 brain Magnetic Resonance (MR) images from subjects that either suffered from subjective memory complaints or were diagnosed with NDs. The segmented WMH volumes correlate well with the standard clinical measurement (Fazekas score), and both the qualitative visualization results and quantitative correlation scores of the proposed algorithm outperform other well known methods for WMH segmentation.

AB - Precise detection and quantification of white matter hyperintensities (WMH) is of great interest in studies of neurodegenerative diseases (NDs). In this work, we propose a novel semi-supervised large margin algorithm for the segmentation of WMH. The proposed algorithm optimizes a kernel based max-margin objective function which aims to maximize the margin averaged over inliers and outliers while exploiting a limited amount of available labelled data. We show that the learning problem can be formulated as a joint framework learning a classifier and a label assignment simultaneously, which can be solved efficiently by an iterative algorithm. We evaluate our method on a database of 280 brain Magnetic Resonance (MR) images from subjects that either suffered from subjective memory complaints or were diagnosed with NDs. The segmented WMH volumes correlate well with the standard clinical measurement (Fazekas score), and both the qualitative visualization results and quantitative correlation scores of the proposed algorithm outperform other well known methods for WMH segmentation.

UR - http://www.scopus.com/inward/record.url?scp=84992484406&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-47157-0_13

DO - 10.1007/978-3-319-47157-0_13

M3 - Paper

SP - 104

EP - 112

ER -

Qin C, Moreno RG, Bowles C, Ledig C, Scheltens P, Barkhof F et al. A semi-supervised large margin algorithm for white matter hyperintensity segmentation. 2016. Paper presented at 7th International Workshop on Machine Learning in Medical Imaging, MLMI 2016 held in conjunction with 19th International Conference on Medical Image Computing and Computer-Assisted Intervention, MICCAI 2016, Athens, Greece. https://doi.org/10.1007/978-3-319-47157-0_13