DuFal AI model enhances sparse-view CT scans by learning dual frequencies

Researchers have developed DuFal, a novel framework for reconstructing high-fidelity Computed Tomography (CT) volumes from extremely limited X-ray projections. This dual-path architecture integrates frequency-domain and spatial-domain processing, utilizing specialized Fourier Neural Operators to capture both global and local high-frequency patterns. The system demonstrated superior performance in preserving fine anatomical details compared to existing methods on the LUNA16 and ToothFairy datasets, particularly in sparse-view scenarios. AI

Summary written by gemini-2.5-flash-lite from 1 source. How we write summaries →

IMPACT Introduces a novel approach to medical image reconstruction, potentially improving diagnostic accuracy from limited scan data.

RANK_REASON This is a research paper detailing a new framework for medical imaging reconstruction. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• DuFal
• LUNA16
• ToothFairy
• arXiv
• Fourier Neural Operator

• paper
• other