Image Restoration for Low-Dose CT via Transfer Learning and Residual Network
Deep learning has recently been extensively investigated to remove artifacts in low-dose computed tomography (LDCT). However, the power of transfer learning for medical image denoising tasks has not been fully explored. In this work, we proposed a transfer learning residual convolutional neural netw...
Main Authors: | , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
IEEE
2020-01-01
|
Series: | IEEE Access |
Subjects: | |
Online Access: | https://ieeexplore.ieee.org/document/9117103/ |
id |
doaj-13b177a37605419ab0cee068222d3d65 |
---|---|
record_format |
Article |
spelling |
doaj-13b177a37605419ab0cee068222d3d652021-03-30T02:26:28ZengIEEEIEEE Access2169-35362020-01-01811207811209110.1109/ACCESS.2020.30025349117103Image Restoration for Low-Dose CT via Transfer Learning and Residual NetworkAnni Zhong0https://orcid.org/0000-0001-5990-2269Bin Li1https://orcid.org/0000-0002-2413-5175Ning Luo2https://orcid.org/0000-0002-6616-2219Yuan Xu3https://orcid.org/0000-0002-2308-3748Linghong Zhou4https://orcid.org/0000-0002-8372-5554Xin Zhen5https://orcid.org/0000-0002-5037-2801School of Biomedical Engineering, Southern Medical University, Guangzhou, ChinaDepartment of Radiation Oncology, State Key Laboratory of Oncology in South China, Collaborative Innovation Center for Cancer Medicine, Sun Yat-sen University Cancer Center, Guangzhou, ChinaDepartment of Radiation Oncology, The First Affiliated Hospital, Sun Yat-sen University, Guangzhou, ChinaSchool of Biomedical Engineering, Southern Medical University, Guangzhou, ChinaSchool of Biomedical Engineering, Southern Medical University, Guangzhou, ChinaSchool of Biomedical Engineering, Southern Medical University, Guangzhou, ChinaDeep learning has recently been extensively investigated to remove artifacts in low-dose computed tomography (LDCT). However, the power of transfer learning for medical image denoising tasks has not been fully explored. In this work, we proposed a transfer learning residual convolutional neural network (TLR-CNN) to restore LDCT images at single and blind noise levels. A residual network was implemented to effectively estimate the difference between denoised image and its original map, and a noise-free image was obtained by subtracting the residual map from the LDCT image. The results were compared to competing baseline denoising methods in terms of quantitative metrics including the PSNR, RMSE, SSIM and FSIM. For the single noise level, the proposed method demonstrated better denoising performance than the other algorithms for both simulation data and clinical data. As for the blind denoising, the image qualities were improved for all noise levels for all the quantitative metrics, but such improvements were decreasing as the noise level decrease (higher mAs). Comparative experiments suggested that the proposed network could effectively suppress artifacts and preserve image details with faster converge rate and reduced computational time.https://ieeexplore.ieee.org/document/9117103/LDCTimage denoisingCNNtransfer learningresidual network |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Anni Zhong Bin Li Ning Luo Yuan Xu Linghong Zhou Xin Zhen |
spellingShingle |
Anni Zhong Bin Li Ning Luo Yuan Xu Linghong Zhou Xin Zhen Image Restoration for Low-Dose CT via Transfer Learning and Residual Network IEEE Access LDCT image denoising CNN transfer learning residual network |
author_facet |
Anni Zhong Bin Li Ning Luo Yuan Xu Linghong Zhou Xin Zhen |
author_sort |
Anni Zhong |
title |
Image Restoration for Low-Dose CT via Transfer Learning and Residual Network |
title_short |
Image Restoration for Low-Dose CT via Transfer Learning and Residual Network |
title_full |
Image Restoration for Low-Dose CT via Transfer Learning and Residual Network |
title_fullStr |
Image Restoration for Low-Dose CT via Transfer Learning and Residual Network |
title_full_unstemmed |
Image Restoration for Low-Dose CT via Transfer Learning and Residual Network |
title_sort |
image restoration for low-dose ct via transfer learning and residual network |
publisher |
IEEE |
series |
IEEE Access |
issn |
2169-3536 |
publishDate |
2020-01-01 |
description |
Deep learning has recently been extensively investigated to remove artifacts in low-dose computed tomography (LDCT). However, the power of transfer learning for medical image denoising tasks has not been fully explored. In this work, we proposed a transfer learning residual convolutional neural network (TLR-CNN) to restore LDCT images at single and blind noise levels. A residual network was implemented to effectively estimate the difference between denoised image and its original map, and a noise-free image was obtained by subtracting the residual map from the LDCT image. The results were compared to competing baseline denoising methods in terms of quantitative metrics including the PSNR, RMSE, SSIM and FSIM. For the single noise level, the proposed method demonstrated better denoising performance than the other algorithms for both simulation data and clinical data. As for the blind denoising, the image qualities were improved for all noise levels for all the quantitative metrics, but such improvements were decreasing as the noise level decrease (higher mAs). Comparative experiments suggested that the proposed network could effectively suppress artifacts and preserve image details with faster converge rate and reduced computational time. |
topic |
LDCT image denoising CNN transfer learning residual network |
url |
https://ieeexplore.ieee.org/document/9117103/ |
work_keys_str_mv |
AT annizhong imagerestorationforlowdosectviatransferlearningandresidualnetwork AT binli imagerestorationforlowdosectviatransferlearningandresidualnetwork AT ningluo imagerestorationforlowdosectviatransferlearningandresidualnetwork AT yuanxu imagerestorationforlowdosectviatransferlearningandresidualnetwork AT linghongzhou imagerestorationforlowdosectviatransferlearningandresidualnetwork AT xinzhen imagerestorationforlowdosectviatransferlearningandresidualnetwork |
_version_ |
1724185151146033152 |