High-resolution CT Image Retrieval Using Sparse Convolutional Neural Network

• Type of Material: Article
• Authors: Yang Lei, Emory University; Dong Xu, Zhejiang Cancer Hospital; Zhengyang Zhou, Emory University; Kristin Higgins, Emory University; Xueyuan Dong, Emory University; Tian Liu, Emory University; Hyunsuk Shim, Emory University; Hui Mao, Emory University; Walter J Curran, Emory University; Xiaofeng Yang, Emory University
• Language: English
• Date: 2018-03-09
• Publisher: Society of Photo-optical Instrumentation Engineers (SPIE)
• Publication Version: Final Published Version
• Copyright Statement: © (2018) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
• Final Published Version (URL): http://dx.doi.org/10.1117/12.2292891
• Title of Journal or Parent Work: Proceedings of SPIE
• ISSN: 0277-786X
• Volume: 10573
• Grant/Funding Information: This research is supported in part by the National Cancer Institute of the National Institutes of Health under Award Number R01CA215718; the Department of Defense (DoD) Prostate Cancer Research Program (PCRP) Award W81XWH-13-1-0269 ;and Dunwoody Golf Club Prostate Cancer Research Award, a philanthropic award provided by the Winship Cancer Institute of Emory University.
• Keywords: high-resolution image retrieval; CT; convolutional neural network

Abstract

We propose a high-resolution CT image retrieval method based on sparse convolutional neural network. The proposed framework is used to train the end-to-end mapping from low-resolution to high-resolution images. The patch-wise feature of low-resolution CT is extracted and sparsely represented by a convolutional layer and a learned iterative shrinkage threshold framework, respectively. Restricted linear unit is utilized to non-linearly map the low-resolution sparse coefficients to the high-resolution ones. An adaptive high-resolution dictionary is applied to construct the informative signature which is highly connected to a high-resolution patch. Finally, we feed the signature to a convolutional layer to reconstruct the predicted high-resolution patches and average these overlapping patches to generate high-resolution CT. The loss function between reconstructed images and the corresponding ground truth high-resolution images is applied to optimize the parameters of end-to-end neural network. The well-trained map is used to generate the high-resolution CT from a new low-resolution input. This technique was tested with brain and lung CT images and the image quality was assessed using the corresponding CT images. Peak signal-to-noise ratio (PSNR), structural similarity index (SSIM) and mean absolute error (MAE) indexes were used to quantify the differences between the generated high-resolution and corresponding ground truth CT images. The experimental results showed the proposed method could enhance images resolution from low-resolution images. The proposed method has great potential in improving radiation dose calculation and delivery accuracy and decreasing CT radiation exposure of patients.