Share Email Print

Proceedings Paper

An unsupervised deep learning approach for 4DCT lung deformable image registration
Author(s): Yabo Fu; Yang Lei; Tonghe Wang; Kristin Higgins; Jeffrey D. Bradley; Walter J. Curran; Tian Liu; Xiaofeng Yang
Format Member Price Non-Member Price
PDF $17.00 $21.00

Paper Abstract

Traditional deformable image registration (DIR) algorithms such as optical flow and demons are iterative and slow especially for large 4D-CT datasets. In order to quickly register the 4D-CT lung images for treatment planning and target definition, the computational speed of the current DIRs needs to be improved. Deep learning-based DIR methods that enable direct transformation prediction are promising alternatives for 4D-CT DIR. In this study, we propose to integrate dilated inception module (DIM) model and self-attention gate (Self-AG) into deep learning framework for 4DCT lung DIR. To overcome the shortage of manually aligned ‘ground truth’ training datasets, the network was designed to train in an unsupervised manner. Instead of using only the fixed and moving images as input, we also included the gradient images in x, y, z directions of the fixed and moving images as input to provide the network additional information to help the transformation prediction. The DIM was able to extract multi-scale structural features for robust feature learning. Self-AG were applied at different scales throughout the encoding and decoding pathways to highlight the structure representing feature differences between moving image and fixed image. The network was trained using pairs of 3D image patches that were extracted from any two random phases of one 4D-CT images. The loss function of the proposed network contains three parts which are image similarity loss, adversarial loss and a regularization loss. The network was trained and tested on 25 patients’ 4D-CT datasets using five-fold out cross validation. The proposed method was evaluated using Mean absolute error (MAE), peak signal to noise ratio (PSNR) and normalized cross correlation (NCC) between the deformed image and the fixed image. MAE, PSNR and NCC were 19.2±6.5, 35.4±3.0 and 0.995±0.002 respectively. Target registration errors (TREs) were calculated using manually selected landmark pairs. The average TRE was 3.38 ± 2.36 mm, which was comparable to traditional DIR algorithms. To summarize, the proposed method was able to achieve comparable performance to that of the traditional DIRs while being orders of magnitude faster (less than a minute).

Paper Details

Date Published: 10 March 2020
PDF: 6 pages
Proc. SPIE 11313, Medical Imaging 2020: Image Processing, 113132T (10 March 2020); doi: 10.1117/12.2549031
Show Author Affiliations
Yabo Fu, Emory Univ. (United States)
Yang Lei, Emory Univ. (United States)
Tonghe Wang, Emory Univ. (United States)
Kristin Higgins, Emory Univ. (United States)
Jeffrey D. Bradley, Emory Univ. (United States)
Walter J. Curran, Emory Univ. (United States)
Tian Liu, Emory Univ. (United States)
Xiaofeng Yang, Emory Univ. (United States)

Published in SPIE Proceedings Vol. 11313:
Medical Imaging 2020: Image Processing
Ivana Išgum; Bennett A. Landman, Editor(s)

© SPIE. Terms of Use
Back to Top
Sign in to read the full article
Create a free SPIE account to get access to
premium articles and original research
Forgot your username?