Scholarlink

Using deep learning to reconstruct CT from X-ray images not only overcomes the inherent 2D limitations of X-ray imaging but also reduces radiation exposure and costs associated with CT scans. While current deep learning approaches offer feasible solutions, they often overlook the role of attention mechanisms in capturing critical anatomical features. Moreover, these methods ignore the perceptual consistency between the reconstructed CT images and the ground truth in terms of luminance, contrast, and structure. To address these limitations, we propose a novel medical image reconstruction framework, namely OX2CT-GAN, which aims to reconstruct 3D chest CT images from 2D orthogonal biplanar X-ray images. Specifically, we propose Feature Enhancement Blocks (FEB), which are embedded in the encoding stage of the generator to enhance the model’s ability to extract critical anatomical features. Additionally, we propose a Perceptual Consistency Loss to enhance the quality of the reconstructed CT images. Experimental results demonstrate that our method significantly outperforms existing deep learning models in both quantitative metrics and qualitative evaluations, achieving state-of-the-art (SOTA) performance with promising applications in the field.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Reconstructing Chest CT from Orthogonal Biplanar X-rays via Feature Enhancement Blocks and Perceptual Consistency Loss

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Tao Tang,
  • Wenjie Cai,
  • Bin Yang,
  • Jun Huang,
  • Ying Zhou,
  • Zhiqin Liu,
  • Qingfeng Wang,
  • Guangwei Li

摘要

Using deep learning to reconstruct CT from X-ray images not only overcomes the inherent 2D limitations of X-ray imaging but also reduces radiation exposure and costs associated with CT scans. While current deep learning approaches offer feasible solutions, they often overlook the role of attention mechanisms in capturing critical anatomical features. Moreover, these methods ignore the perceptual consistency between the reconstructed CT images and the ground truth in terms of luminance, contrast, and structure. To address these limitations, we propose a novel medical image reconstruction framework, namely OX2CT-GAN, which aims to reconstruct 3D chest CT images from 2D orthogonal biplanar X-ray images. Specifically, we propose Feature Enhancement Blocks (FEB), which are embedded in the encoding stage of the generator to enhance the model’s ability to extract critical anatomical features. Additionally, we propose a Perceptual Consistency Loss to enhance the quality of the reconstructed CT images. Experimental results demonstrate that our method significantly outperforms existing deep learning models in both quantitative metrics and qualitative evaluations, achieving state-of-the-art (SOTA) performance with promising applications in the field.