Semi-supervised learning for pelvic MR image segmentation based on multi-task residual fully convolutional networks

Zishun Feng, Dong Nie, Li Wang, Dinggang Shen

Research output: Chapter in Book/Report/Conference proceedingConference contribution

5 Citations (Scopus)

Abstract

Accurate segmentation of pelvic organs from magnetic resonance (MR) images plays an important role in image-guided radiotherapy. However, it is a challenging task due to inconsistent organ appearances and large shape variations. Fully convolutional network (FCN) has recently achieved state-of-the-art performance in medical image segmentation, but it requires a large amount of labeled data for training, which is usually difficult to obtain in real situation. To address these challenges, we propose a deep learning based semi-supervised learning framework. Specifically, we first train an initial multi-task residual fully convolutional network (FCN) based on a limited number of labeled MRI data. Based on the initially trained FCN, those unlabeled new data can be automatically segmented and some reasonable segmentations (after manual/automatic checking) can be included into the training data to fine-tune the network. This step can be repeated to progressively improve the training of our network, until no reasonable segmentations of new data can be included. Experimental results demonstrate the effectiveness of our proposed progressive semi-supervised learning fashion as well as its advantage in terms of accuracy.

Original languageEnglish
Title of host publication2018 IEEE 15th International Symposium on Biomedical Imaging, ISBI 2018
PublisherIEEE Computer Society
Pages885-888
Number of pages4
Volume2018-April
ISBN (Electronic)9781538636367
DOIs
Publication statusPublished - 2018 May 23
Externally publishedYes
Event15th IEEE International Symposium on Biomedical Imaging, ISBI 2018 - Washington, United States
Duration: 2018 Apr 42018 Apr 7

Other

Other15th IEEE International Symposium on Biomedical Imaging, ISBI 2018
CountryUnited States
CityWashington
Period18/4/418/4/7

Fingerprint

Supervised learning
Magnetic resonance
Image segmentation
Magnetic Resonance Spectroscopy
Image-Guided Radiotherapy
Radiotherapy
Magnetic resonance imaging
Learning
Supervised Machine Learning
Deep learning

Keywords

  • Fully convolutional network
  • Multi-task learning
  • Neural network
  • Pelvic MRI segmentation
  • Semi-supervised learning

ASJC Scopus subject areas

  • Biomedical Engineering
  • Radiology Nuclear Medicine and imaging

Cite this

Feng, Z., Nie, D., Wang, L., & Shen, D. (2018). Semi-supervised learning for pelvic MR image segmentation based on multi-task residual fully convolutional networks. In 2018 IEEE 15th International Symposium on Biomedical Imaging, ISBI 2018 (Vol. 2018-April, pp. 885-888). IEEE Computer Society. https://doi.org/10.1109/ISBI.2018.8363713

Semi-supervised learning for pelvic MR image segmentation based on multi-task residual fully convolutional networks. / Feng, Zishun; Nie, Dong; Wang, Li; Shen, Dinggang.

2018 IEEE 15th International Symposium on Biomedical Imaging, ISBI 2018. Vol. 2018-April IEEE Computer Society, 2018. p. 885-888.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Feng, Z, Nie, D, Wang, L & Shen, D 2018, Semi-supervised learning for pelvic MR image segmentation based on multi-task residual fully convolutional networks. in 2018 IEEE 15th International Symposium on Biomedical Imaging, ISBI 2018. vol. 2018-April, IEEE Computer Society, pp. 885-888, 15th IEEE International Symposium on Biomedical Imaging, ISBI 2018, Washington, United States, 18/4/4. https://doi.org/10.1109/ISBI.2018.8363713
Feng Z, Nie D, Wang L, Shen D. Semi-supervised learning for pelvic MR image segmentation based on multi-task residual fully convolutional networks. In 2018 IEEE 15th International Symposium on Biomedical Imaging, ISBI 2018. Vol. 2018-April. IEEE Computer Society. 2018. p. 885-888 https://doi.org/10.1109/ISBI.2018.8363713
Feng, Zishun ; Nie, Dong ; Wang, Li ; Shen, Dinggang. / Semi-supervised learning for pelvic MR image segmentation based on multi-task residual fully convolutional networks. 2018 IEEE 15th International Symposium on Biomedical Imaging, ISBI 2018. Vol. 2018-April IEEE Computer Society, 2018. pp. 885-888
@inproceedings{49306044d38840dc91e2246d7f0f895e,
title = "Semi-supervised learning for pelvic MR image segmentation based on multi-task residual fully convolutional networks",
abstract = "Accurate segmentation of pelvic organs from magnetic resonance (MR) images plays an important role in image-guided radiotherapy. However, it is a challenging task due to inconsistent organ appearances and large shape variations. Fully convolutional network (FCN) has recently achieved state-of-the-art performance in medical image segmentation, but it requires a large amount of labeled data for training, which is usually difficult to obtain in real situation. To address these challenges, we propose a deep learning based semi-supervised learning framework. Specifically, we first train an initial multi-task residual fully convolutional network (FCN) based on a limited number of labeled MRI data. Based on the initially trained FCN, those unlabeled new data can be automatically segmented and some reasonable segmentations (after manual/automatic checking) can be included into the training data to fine-tune the network. This step can be repeated to progressively improve the training of our network, until no reasonable segmentations of new data can be included. Experimental results demonstrate the effectiveness of our proposed progressive semi-supervised learning fashion as well as its advantage in terms of accuracy.",
keywords = "Fully convolutional network, Multi-task learning, Neural network, Pelvic MRI segmentation, Semi-supervised learning",
author = "Zishun Feng and Dong Nie and Li Wang and Dinggang Shen",
year = "2018",
month = "5",
day = "23",
doi = "10.1109/ISBI.2018.8363713",
language = "English",
volume = "2018-April",
pages = "885--888",
booktitle = "2018 IEEE 15th International Symposium on Biomedical Imaging, ISBI 2018",
publisher = "IEEE Computer Society",

}

TY - GEN

T1 - Semi-supervised learning for pelvic MR image segmentation based on multi-task residual fully convolutional networks

AU - Feng, Zishun

AU - Nie, Dong

AU - Wang, Li

AU - Shen, Dinggang

PY - 2018/5/23

Y1 - 2018/5/23

N2 - Accurate segmentation of pelvic organs from magnetic resonance (MR) images plays an important role in image-guided radiotherapy. However, it is a challenging task due to inconsistent organ appearances and large shape variations. Fully convolutional network (FCN) has recently achieved state-of-the-art performance in medical image segmentation, but it requires a large amount of labeled data for training, which is usually difficult to obtain in real situation. To address these challenges, we propose a deep learning based semi-supervised learning framework. Specifically, we first train an initial multi-task residual fully convolutional network (FCN) based on a limited number of labeled MRI data. Based on the initially trained FCN, those unlabeled new data can be automatically segmented and some reasonable segmentations (after manual/automatic checking) can be included into the training data to fine-tune the network. This step can be repeated to progressively improve the training of our network, until no reasonable segmentations of new data can be included. Experimental results demonstrate the effectiveness of our proposed progressive semi-supervised learning fashion as well as its advantage in terms of accuracy.

AB - Accurate segmentation of pelvic organs from magnetic resonance (MR) images plays an important role in image-guided radiotherapy. However, it is a challenging task due to inconsistent organ appearances and large shape variations. Fully convolutional network (FCN) has recently achieved state-of-the-art performance in medical image segmentation, but it requires a large amount of labeled data for training, which is usually difficult to obtain in real situation. To address these challenges, we propose a deep learning based semi-supervised learning framework. Specifically, we first train an initial multi-task residual fully convolutional network (FCN) based on a limited number of labeled MRI data. Based on the initially trained FCN, those unlabeled new data can be automatically segmented and some reasonable segmentations (after manual/automatic checking) can be included into the training data to fine-tune the network. This step can be repeated to progressively improve the training of our network, until no reasonable segmentations of new data can be included. Experimental results demonstrate the effectiveness of our proposed progressive semi-supervised learning fashion as well as its advantage in terms of accuracy.

KW - Fully convolutional network

KW - Multi-task learning

KW - Neural network

KW - Pelvic MRI segmentation

KW - Semi-supervised learning

UR - http://www.scopus.com/inward/record.url?scp=85048084319&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85048084319&partnerID=8YFLogxK

U2 - 10.1109/ISBI.2018.8363713

DO - 10.1109/ISBI.2018.8363713

M3 - Conference contribution

AN - SCOPUS:85048084319

VL - 2018-April

SP - 885

EP - 888

BT - 2018 IEEE 15th International Symposium on Biomedical Imaging, ISBI 2018

PB - IEEE Computer Society

ER -