Reproduce “UncertaINR: Uncertainty Quantification of End-to-End Implicit Neural Representations for Computed Tomography” 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
Francisca Vasconcelos, Bobby He, Nalini Singh, Yee Whye Teh
UncertaINR reconstructs CT images from X-ray projections and estimates pixel-wise uncertainty. It represents the image with a coordinate-based network, fits predictions to measured projections, and uses stochastic dropout to produce reconstructions and uncertainty maps. On eight simulated AAPM test images at 60 views, adding dropout to a 10-model ensemble lowered adjusted ECE from 0.144 to 0.045; this does not establish clinical benefit.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.