Dongsheng Jiang1, Weiqiang Dou2,3, Luc Vosters4, Xiayu Xu5,6, Yue Sun4, Tao Tan7,8. 1. School of Basic Medical Science, Digital Medical Research Center, Fudan University, Shanghai, People's Republic of China. 2. Department of Radiology and Nuclear Medicine, Radboud University Medical Center, Nijmegen, The Netherlands. 3. GE Healthcare, MR Research China, Beijing, People's Republic of China. 4. Department of Electrical Engineering, Eindhoven University of Technology, Eindhoven, The Netherlands. 5. The Key Laboratory of Biomedical Information Engineering of Ministry of Education, School of Life Science and Technology, Xi'an Jiaotong University, Xi'an, People's Republic of China. 6. Bioinspired Engineering and Biomechanics Center (BEBC), Xi'an Jiaotong University, Xi'an, People's Republic of China. 7. Department of Biomedical Engineering, Eindhoven University of Technology, Eindhoven, The Netherlands. tao.tan911@gmail.com. 8. ScreenPoint Medical, Nijmegen, The Netherlands. tao.tan911@gmail.com.
Abstract
PURPOSE: To test if the proposed deep learning based denoising method denoising convolutional neural networks (DnCNN) with residual learning and multi-channel strategy can denoise three dimensional MR images with Rician noise robustly. MATERIALS AND METHODS: Multi-channel DnCNN (MCDnCNN) method with two training strategies was developed to denoise MR images with and without a specific noise level, respectively. To evaluate our method, three datasets from two public data sources of IXI dataset and Brainweb, including T1 weighted MR images acquired at 1.5 and 3 T as well as MR images simulated with a widely used MR simulator, were randomly selected and artificially added with different noise levels ranging from 1 to 15%. For comparison, four other state-of-the-art denoising methods were also tested using these datasets. RESULTS: In terms of the highest peak-signal-to-noise-ratio and global of structure similarity index, our proposed MCDnCNN model for a specific noise level showed the most robust denoising performance in all three datasets. Next to that, our general noise-applicable model also performed better than the rest four methods in two datasets. Furthermore, our training model showed good general applicability. CONCLUSION: Our proposed MCDnCNN model has been demonstrated to robustly denoise three dimensional MR images with Rician noise.
PURPOSE: To test if the proposed deep learning based denoising method denoising convolutional neural networks (DnCNN) with residual learning and multi-channel strategy can denoise three dimensional MR images with Rician noise robustly. MATERIALS AND METHODS: Multi-channel DnCNN (MCDnCNN) method with two training strategies was developed to denoise MR images with and without a specific noise level, respectively. To evaluate our method, three datasets from two public data sources of IXI dataset and Brainweb, including T1 weighted MR images acquired at 1.5 and 3 T as well as MR images simulated with a widely used MR simulator, were randomly selected and artificially added with different noise levels ranging from 1 to 15%. For comparison, four other state-of-the-art denoising methods were also tested using these datasets. RESULTS: In terms of the highest peak-signal-to-noise-ratio and global of structure similarity index, our proposed MCDnCNN model for a specific noise level showed the most robust denoising performance in all three datasets. Next to that, our general noise-applicable model also performed better than the rest four methods in two datasets. Furthermore, our training model showed good general applicability. CONCLUSION: Our proposed MCDnCNN model has been demonstrated to robustly denoise three dimensional MR images with Rician noise.
Keywords:
CNN; Deep learning; Denoising; MRI; Rician noise
Authors: Albert Juan Ramon; Yongyi Yang; P Hendrik Pretorius; Karen L Johnson; Michael A King; Miles N Wernick Journal: IEEE Trans Med Imaging Date: 2020-03-10 Impact factor: 11.037