Novel-view X-ray projection synthesis through geometry-integrated deep learning

Liyue Shen,Lequan Yu,Wei Zhao,John Pauly,Lei Xing

DOI: https://doi.org/10.1016/j.media.2022.102372

IF: 10.9

2022-04-01

Medical Image Analysis

Abstract:X-ray imaging is a widely used approach to view the internal structure of a subject for clinical diagnosis, image-guided interventions and decision-making. The X-ray projections acquired at different view angles provide complementary information of patient's anatomy and are required for stereoscopic or volumetric imaging of the subject. In reality, obtaining multiple-view projections inevitably increases radiation dose and complicates clinical workflow. Here we investigate a strategy of obtaining the X-ray projection image at a novel view angle from a given projection image at a specific view angle to alleviate the need for actual projection measurement. Specifically, a Deep Learning-based Geometry-Integrated Projection Synthesis (DL-GIPS) framework is proposed for the generation of novel-view X-ray projections. The proposed deep learning model extracts geometry and texture features from a source-view projection, and then conducts geometry transformation on the geometry features to accommodate the change of view angle. At the final stage, the X-ray projection in the target view is synthesized from the transformed geometry and the shared texture features via an image generator. The feasibility and potential impact of the proposed DL-GIPS model are demonstrated using lung imaging cases. The proposed strategy can be generalized to a general case of multiple projections synthesis from multiple input views and potentially provides a new paradigm for various stereoscopic and volumetric imaging with substantially reduced efforts in data acquisition.

engineering, biomedical,computer science, interdisciplinary applications, artificial intelligence,radiology, nuclear medicine & medical imaging