Share Email Print
cover

Proceedings Paper • new

Deep learning for low-dose CT
Author(s): Hu Chen; Yi Zhang; Jiliu Zhou; Ge Wang
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Given the potential risk of X-ray radiation to the patient, low-dose CT has attracted a considerable interest in the medical imaging field. Currently, the main stream low-dose CT methods include vendor-specific sinogram domain filtration and iterative reconstruction algorithms, but they need to access raw data whose formats are not transparent to most users. Due to the difficulty of modeling the statistical characteristics in the image domain, the existing methods for directly processing reconstructed images cannot eliminate image noise very well while keeping structural details. Inspired by the idea of deep learning, here we combine the autoencoder, deconvolution network, and shortcut connections into the residual encoder-decoder convolutional neural network (RED-CNN) for low-dose CT imaging. After patch-based training, the proposed RED-CNN achieves a competitive performance relative to the-state-of-art methods. Especially, our method has been favorably evaluated in terms of noise suppression and structural preservation.

Paper Details

Date Published: 19 September 2017
PDF: 8 pages
Proc. SPIE 10391, Developments in X-Ray Tomography XI, 103910I (19 September 2017); doi: 10.1117/12.2272723
Show Author Affiliations
Hu Chen, Sichuan Univ. (China)
Yi Zhang, Sichuan Univ. (China)
Jiliu Zhou, Sichuan Univ. (China)
Ge Wang, Rensselaer Polytechnic Institute (United States)


Published in SPIE Proceedings Vol. 10391:
Developments in X-Ray Tomography XI
Bert Müller; Ge Wang, Editor(s)

© SPIE. Terms of Use
Back to Top