New AI model learns 2D-3D bone correspondence for X-ray to CT registration

Researchers have developed a novel method for registering knee bone poses from X-ray images to pre-operative CT scans. This approach learns a dense 2D-3D correspondence across 758 patients, enabling subject-agnostic registration without requiring individual patient CT data or iterative rendering. The learned representation is also anatomically semantic, allowing for landmark classification and bone segmentation without explicit labels. AI

IMPACT This research could improve the accuracy and efficiency of medical imaging analysis by enabling better alignment of different imaging modalities.

RANK_REASON The cluster contains a research paper detailing a new AI model for medical image registration. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• 758 patients
• alphaXiv
• CatalyzeX
• computed tomography
• CORE Recommender
• DagsHub
• Gotit.pub
• Hugging Face
• Influence Flower
• Knee Bones
• radiography
• ScienceCast
• X-ray

• paper
• other

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →