New CAN-FLOW framework generates realistic cardiac anatomy for virtual cohorts

Researchers have developed CAN-FLOW, a novel framework for generating realistic cardiac anatomy data for virtual cohorts. This method utilizes conditional normalizing flows to model anatomical variability based on factors like sex, age, and body mass index. Trained on data from 2,208 UK Biobank subjects, CAN-FLOW demonstrated superior performance over conditional variational autoencoders in reproducing clinical phenotype distributions and anatomical trends. The framework aims to address data limitations in cardiac digital twin research and facilitate in silico clinical trial workflows. AI

IMPACT This research could accelerate the development of digital twins for medical research by providing a method to generate diverse and realistic anatomical data.

RANK_REASON The cluster contains an academic paper detailing a new method for generating synthetic data. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.LG →

• arXiv
• Hugging Face
• Mathias Peirlinck
• normalizing flows
• UK Biobank

• paper
• model release

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