Comparative clinical evaluation of atlas and deep-learning-based auto-segmentation of organ structures in liver cancer

被引:66
作者
Ahn, Sang Hee [1 ]
Yeo, Adam Unjin [2 ]
Kim, Kwang Hyeon [1 ]
Kim, Chankyu [1 ]
Goh, Youngmoon [3 ]
Cho, Shinhaeng [4 ]
Lee, Se Byeong [1 ]
Lim, Young Kyung [1 ]
Kim, Haksoo [1 ]
Shin, Dongho [1 ]
Kim, Taeyoon [1 ]
Kim, Tae Hyun [1 ]
Youn, Sang Hee [1 ]
Oh, Eun Sang [1 ]
Jeong, Jong Hwi [1 ]
机构
[1] Natl Canc Ctr, Proton Therapy Ctr, Dept Radiat Oncol, 323 Ilsan Ro, Goyang Si 10408, Gyeonggi Do, South Korea
[2] Peter MacCallum Canc Ctr, Melbourne, Vic, Australia
[3] Asan Med Ctr, Dept Radiat Oncol, Seoul, South Korea
[4] Chonnam Natl Univ, Dept Radiat Oncol, Med Sch, Gwangju, South Korea
关键词
Contouring; Atlas-based auto-segmentation; Deep-learning-based auto-segmentation; Deep convolution neural network (DCNN); CT IMAGES; REGISTRATION; HEAD;
D O I
10.1186/s13014-019-1392-z
中图分类号
R73 [肿瘤学];
学科分类号
100214 ;
摘要
Background Accurate and standardized descriptions of organs at risk (OARs) are essential in radiation therapy for treatment planning and evaluation. Traditionally, physicians have contoured patient images manually, which, is time-consuming and subject to inter-observer variability. This study aims to a) investigate whether customized, deep-learning-based auto-segmentation could overcome the limitations of manual contouring and b) compare its performance against a typical, atlas-based auto-segmentation method organ structures in liver cancer. Methods On-contrast computer tomography image sets of 70 liver cancer patients were used, and four OARs (heart, liver, kidney, and stomach) were manually delineated by three experienced physicians as reference structures. Atlas and deep learning auto-segmentations were respectively performed with MIM Maestro 6.5 (MIM Software Inc., Cleveland, OH) and, with a deep convolution neural network (DCNN). The Hausdorff distance (HD) and, dice similarity coefficient (DSC), volume overlap error (VOE), and relative volume difference (RVD) were used to quantitatively evaluate the four different methods in the case of the reference set of the four OAR structures. Results The atlas-based method yielded the following average DSC and standard deviation values (SD) for the heart, liver, right kidney, left kidney, and stomach: 0.92 +/- 0.04 (DSC +/- SD), 0.93 +/- 0.02, 0.86 +/- 0.07, 0.85 +/- 0.11, and 0.60 +/- 0.13 respectively. The deep-learning-based method yielded corresponding values for the OARs of 0.94 +/- 0.01, 0.93 +/- 0.01, 0.88 +/- 0.03, 0.86 +/- 0.03, and 0.73 +/- 0.09. The segmentation results show that the deep learning framework is superior to the atlas-based framwork except in the case of the liver. Specifically, in the case of the stomach, the DSC, VOE, and RVD showed a maximum difference of 21.67, 25.11, 28.80% respectively. Conclusions In this study, we demonstrated that a deep learning framework could be used more effectively and efficiently compared to atlas-based auto-segmentation for most OARs in human liver cancer. Extended use of the deep-learning-based framework is anticipated for auto-segmentations of other body sites.
引用
收藏
页码:1 / 13
页数:13
相关论文
共 38 条
  • [1] Social-Sensor Cloud Service Selection
    Aamir, Tooba
    Bouguettaya, Athman
    Dong, Hai
    Erradi, Abdelkarim
    Hadjidj, Rachid
    [J]. 2017 IEEE 24TH INTERNATIONAL CONFERENCE ON WEB SERVICES (ICWS 2017), 2017, : 508 - 515
  • [2] Allaire J. J., 2018, DEEP LEARNING WITH R
  • [3] [Anonymous], 2015, DEEP LEARNING NATURE, DOI [10.1038/nature14539, DOI 10.1038/NATURE14539]
  • [4] Diagnostic Assessment of Deep Learning Algorithms for Detection of Lymph Node Metastases in Women With Breast Cancer
    Bejnordi, Babak Ehteshami
    Veta, Mitko
    van Diest, Paul Johannes
    van Ginneken, Bram
    Karssemeijer, Nico
    Litjens, Geert
    van der Laak, Jeroen A. W. M.
    [J]. JAMA-JOURNAL OF THE AMERICAN MEDICAL ASSOCIATION, 2017, 318 (22): : 2199 - 2210
  • [5] Combining registration and active shape models for the automatic segmentation of the lymph node regions in head and neck CT images
    Chen, Antong
    Deeley, Matthew A.
    Niermann, Kenneth J.
    Moretti, Luigi
    Dawant, Benoit M.
    [J]. MEDICAL PHYSICS, 2010, 37 (12) : 6338 - 6346
  • [6] Christ PF, 2017, ARXIV170205970
  • [7] Atlas-based automatic segmentation of head and neck organs at risk and nodal target volumes: a clinical validation
    Daisne, Jean-Francois
    Blumhofer, Andreas
    [J]. RADIATION ONCOLOGY, 2013, 8
  • [8] Automatic Multi-Organ Segmentation on Abdominal CT With Dense V-Networks
    Gibson, Eli
    Giganti, Francesco
    Hu, Yipeng
    Bonmati, Ester
    Bandula, Steve
    Gurusamy, Kurinchi
    Davidson, Brian
    Pereira, Stephen P.
    Clarkson, Matthew J.
    Barratt, Dean C.
    [J]. IEEE TRANSACTIONS ON MEDICAL IMAGING, 2018, 37 (08) : 1822 - 1834
  • [9] Hinton G.E., 2012, ARXIV PREPRINT ARXIV
  • [10] Automatic 3D liver segmentation based on deep learning and globally optimized surface evolution
    Hu, Peijun
    Wu, Fa
    Peng, Jialin
    Liang, Ping
    Kong, Dexing
    [J]. PHYSICS IN MEDICINE AND BIOLOGY, 2016, 61 (24) : 8676 - 8698