Reproduce “Convex Potential Learning for Optimal Transport Between Medical Images” 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
Hongyu Kan, Ivan Medri, Natasha Ironside, Kristofor Pas, Richard Corp, Shinjini Kundu, Gustavo Kunde Rohde
The authors develop a method for image transport that represents a convex potential with a Log-Sum-Exp function and uses its gradient as the map. They fit it by reducing mass-preservation residuals. In lobar intracerebral hemorrhage CT, it reports L1 error of 0.287 ± 0.139, versus 1.098 ± 0.398 for Kundu et al.; this measures push-forward matching, not map accuracy or downstream benefit.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.