解决胸部X光片多标签与零样本病变分类难题
CXR-LT 2026 Challenge: Projection-Aware Multi-Label and Zero-Shot Chest X-Ray Classification
- 融合投影特异性模型构建统一分类框架
- 零样本任务中实现高泛化性能，克服长尾分布问题
- 适合医疗AI研究者和临床辅助诊断系统开发者
本挑战聚焦于已知胸部X光（CXR）病灶的多标签分类与未见病灶的零样本分类。为应对多种CXR投影视角差异，我们通过分类网络整合投影特异性模型，构建统一框架。针对零样本分类（任务2），在CheXzero基础上引入新型双分支结构，结合对比学习、非对称损失（ASL）及大语言模型生成的描述性提示，有效缓解严重长尾分布问题，并最大化零样本泛化能力。此外，强大的数据增强与测试时增强（TTA）策略确保了两项任务的鲁棒性。
原文摘要 · Abstract (English)
This challenge tackles multi-label classification for known chest X-ray (CXR) lesions and zero-shot classification for unseen ones. To handle diverse CXR projections, we integrate projection-specific models via a classification network into a unified framework. For zero-shot classification (Task 2), we extend CheXzero with a novel dual-branch architecture that combines contrastive learning, Asymmetric Loss (ASL), and LLM-generated descriptive prompts. This effectively mitigates severe long-tail imbalances and maximizes zero-shot generalization. Additionally, strong data and test-time augmentations (TTA) ensure robustness across both tasks.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。