paper-with-me

CheXpert

홈페이지 · 논문 628편

The CheXpert dataset contains 224,316 chest radiographs of 65,240 patients with both frontal and lateral views available. The task is to do automated chest x-ray interpretation, featuring uncertainty labels and radiologist-labeled reference standard evaluation sets. Source: Deep Mining External Imperfect Data for Chest X-ray Disease Screening Image Source: https://stanfordmlgroup.github.io/competitions/chexpert/

ImagesMedical EnglishKorean

벤치마크

Multi-Label Classification on CheXpert 결과 226개
Concept-based Classification on CheXpert 결과 2개