Reproduce “Towards Scalable Diagnostic-Driven Whole-Slide Image Retrieval for Pathology Archives” 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

Jingsong Liu, Youssef Wally, Han Li, Weiwei Zhou, Jing Dai, Elisabeth Wetzer, Peter J. Schüffler

CLEAR-WSI turns each whole-slide image into one fixed-length visual vector, ranks archive cases by cosine similarity, and filters candidates using predicted diagnoses when query labels are unavailable. It adds each processed slide to a growing archive without retraining. On BRACS-1, automated retrieval reaches 71.03% Precision@5 versus RetCCL’s 68.28%; this measures label agreement, not clinical usefulness or hospital-scale speed.

Summary by alphaXiv.

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