A modality-adaptive method for segmenting brain tumors and organs-at-risk in radiation therapy planning

被引：32

作者：

Agn, Mikael ^{[1
]}

af Rosenschold, Per Munck ^{[2
]}

Puonti, Oula ^{[3
]}

Lundemann, Michael J. ^{[4
]}

Mancini, Laura ^{[5
,6
]}

Papadaki, Anastasia ^{[5
,6
]}

Thust, Steffi ^{[5
,6
]}

Ashburner, John ^{[7
]}

Law, Ian ^{[8
]}

Van Leemput, Koen ^{[1
,9
]}

机构：

[1] Tech Univ Denmark, Dept Appl Math & Comp Sci, Lyngby, Denmark

[2] Skane Univ Hosp, Dept Hematol Oncol & Radiat Phys, Radiat Phys, Lund, Sweden

[3] Copenhagen Univ Hosp, Danish Res Ctr Magnet Resonance, Hvidovre, Denmark

[4] Copenhagen Univ Hosp, Rigshosp, Dept Oncol, Copenhagen, Denmark

[5] UCL, UCL Inst Neurol, Dept Brain Repair & Rehabil, Neuroradiol Acad Unit, London, England

[6] UCLH NHS Fdn Trust, Natl Hosp Neurol & Neurosurg, Lysholm Dept Neuroradiol, London, England

[7] UCL, UCL Inst Neurol, Wellcome Ctr Human Neuroimaging, London, England

[8] Copenhagen Univ Hosp, Rigshosp, Dept Clin Physiol Nucl Med & PET, Copenhagen, Denmark

[9] Harvard Med Sch, Massachusetts Gen Hosp, Athinoula A Martinos Ctr Biomed Imaging, Boston, MA 02115 USA

来源：

MEDICAL IMAGE ANALYSIS | 2019年 / 54卷

基金：

英国惠康基金;

关键词：

Glioma; Whole-brain segmentation; Generative probabilistic model; Restricted Boltzmann machine; SEGMENTATION; MODEL; RADIOTHERAPY; ATLAS; MRI; IMAGES; CT; DELINEATION; NETWORKS; FORESTS;

D O I：

10.1016/j.media.2019.03.005

中图分类号：

TP18 [人工智能理论];

学科分类号：

081104 ; 0812 ; 0835 ; 1405 ;

摘要：

In this paper we present a method for simultaneously segmenting brain tumors and an extensive set of organs-at-risk for radiation therapy planning of glioblastomas. The method combines a contrast-adaptive generative model for whole-brain segmentation with a new spatial regularization model of tumor shape using convolutional restricted Boltzmann machines. We demonstrate experimentally that the method is able to adapt to image acquisitions that differ substantially from any available training data, ensuring its applicability across treatment sites; that its tumor segmentation accuracy is comparable to that of the current state of the art; and that it captures most organs-at-risk sufficiently well for radiation therapy planning purposes. The proposed method may be a valuable step towards automating the delineation of brain tumors and organs-at-risk in glioblastoma patients undergoing radiation therapy. (C) 2019 The Authors. Published by Elsevier B.V.

引用

页码：220 / 237

页数：18

共 80 条

[1] Impact of [18F]-fluoro-ethyl-tyrosine PET imaging on target definition for radiation therapy of high-grade glioma
af Rosenschold, Per Munck
Costa, Junia
Engelholm, Svend Aage
Lundemann, Michael J.
Law, Ian
Ohlhues, Lars
Engelholm, Silke
[J]. NEURO-ONCOLOGY, 2015, 17 (05) : 757 - 763
[2] Agn Mikael, 2016, Brainlesion: Glioma, Multiple Sclerosis, Stroke and Traumatic Brain Injuries. First International Workshop, Brainles 2015, held in conjunction with MICCAI 2015. Revised Selected Papers: LNCS 9556, P168, DOI 10.1007/978-3-319-30858-6_15
[3] [Anonymous], P SPIE MED IMAGING
[4] [Anonymous], ARXIV13111354
[5] [Anonymous], INT J RAD ONCOL BIOL
[6] [Anonymous], NEUROIMAGE
[7] [Anonymous], 1992, ADV NEURAL INFORM PR
[8] [Anonymous], ARXIV180600363
[9] [Anonymous], 2012, MACHINE LEARNING PRO
[10] Ashburner J, 2000, HUM BRAIN MAPP, V9, P212, DOI 10.1002/(SICI)1097-0193(200004)9:4<212::AID-HBM3>3.0.CO

← 1 2 3 4 5 6 7 8 →