KEYWORDS: Cone beam computed tomography, Image restoration, Transformers, 3D projection, Deep learning, 3D modeling, Reconstruction algorithms, 3D image reconstruction, 3D image processing, Data modeling
Cone beam CT (CBCT) imaging with sparse-view can effectively reduce the radiation dose risk. The convolution-based end-to-end deep learning methods have been used in single-view CBCT image reconstruction, which can minimize radiation dose and achieve fast CBCT imaging. However, these methods ignore the mismatch between the local feature extraction ability of convolutional neural network (CNN) and the global features of the projection image. To address this issue, we propose a novel deep learning network architecture based on Swin Transformer for single-view CBCT reconstruction. First, we use the Swin Transformer network block to construct a single-view projection feature extraction module, then through the feature transformation module, we convert the 2D features learned from projection into 3D feature tensors, and finally get the 3D volume image in the generative network. This paper is the first attempt to use the Swin Transformer model for single-view CBCT reconstruction. Experimental results with the mouse datasets demonstrate that the proposed model outperforms the convolution-based end-to-end deep learning model in reducing artifacts and preserving the image accuracy.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.