New PRDiT framework generates high-resolution 3D CT volumes

Researchers have introduced the Pixel-Level Residual Diffusion Transformer (PRDiT), a novel framework designed for generating high-resolution 3D CT volumes with intricate details. This two-stage model first uses an MLP-based blind estimator for efficient low-frequency structure separation and then employs a memory-efficient attention transformer to model and refine high-frequency residuals across entire volumes. Experiments on LIDC-IDRI and RAD-ChestCT datasets show PRDiT surpasses existing models like HA-GAN, 3D LDM, and WDM-3D in generating detailed medical imagery. AI

IMPACT Introduces a new method for generating detailed 3D medical imagery, potentially improving diagnostic tools and research.

RANK_REASON The cluster contains a research paper detailing a new generative model for 3D CT volume generation. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CV →

• 3D LDM
• LIDC-IDRI
• multilayer perceptron
• Pixel-Level Residual Diffusion Transformer
• PRDiT
• Rad-ChestCT
• WDM-3D

• paper
• model release

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