Reproduce “A 3D foundation model unifying CT interpretation and segmentation” 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

Yulu Xie, Zilong Li, Shen Xu, Yifan Bai, Liuhong Shi, Jiao Chen, Xianglei Kong, Wenchao Zhu and 9 more

HoloCT combines whole-volume and fine-grained 3D CT features for clinical interpretation and, when requested, segmentation. It sends both feature streams to a language model and uses a mask decoder for voxel-level localization. In retrospective held-out thyroid and colon screening cohorts, fine-tuned HoloCT-8B achieved an equally weighted mean accuracy of 85.1%, exceeding four radiologists; this does not establish general clinical performance.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.