Abstract
Chest X-ray bone suppression is crucial for improving the visibility of soft tissue structures and enhancing diagnostic accuracy. However, traditional approaches often struggle with maintaining high resolution and anatomical accuracy. In this work, we propose a novel approach for high-resolution chest X-ray bone suppression using unpaired CT structural priors.
Our method leverages the rich structural information available in CT scans to guide bone suppression in X-ray images, even without paired training data. The framework consists of three main components: (1) a structural prior extraction module that learns anatomical knowledge from CT scans, (2) an unpaired learning framework that transfers structural knowledge to X-ray domain, and (3) a high-resolution generation module that produces detailed bone-suppressed images.
We evaluate our approach on a large dataset of chest X-ray images with expert annotations. Experimental results demonstrate that our method significantly improves bone suppression quality compared to existing approaches, particularly in high-resolution scenarios. The method maintains anatomical accuracy while providing enhanced soft tissue visibility.
The proposed framework represents a significant advancement in chest X-ray processing, providing higher quality images that could improve diagnostic accuracy and clinical interpretation.
BibTeX
@article{li2020high,
title={High-Resolution Chest X-ray Bone Suppression Using Unpaired CT Structural Priors},
author={Li, Han and Han, Hu and Li, Zeju and Wang, Lei and Wu, Zhe and Lu, Jingjing and Zhou, S Kevin},
journal={IEEE Transactions on Medical Imaging},
year={2020},
publisher={IEEE},
doi={10.1109/TMI.2020.2986242}
}