Reproduce “Conditional Flow Matching for Cross-Field MRI Harmonisation” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Baris Imre, Aram Salehi, Levente Baljer, Andrew Webb, Marius Staring, Efe Ilicak
The model translates brain MRI slices among five field strengths and three contrasts with one shared system for all 60 directed field-and-contrast combinations. It first learns restoration from unpaired scans, then learns field-to-field changes from scarce paired scans using conditional flow matching. Across held-out tests, five Heun steps yield 0.909 SSIM, but voxel metrics do not establish anatomical safety.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.