AlphaRAD：基于α校正二元交叉熵与分解隐层监督的胸部放射学 grounded 零样本分类
AlphaRAD: Grounded Zero-Shot Classification in Chest Radiology via $α$-Corrected Binary Cross Entropy and Factorized Latent Supervision
浏览论文内容
中文总结 AI 辅助
AlphaRAD通过α校正二元交叉熵训练的医疗概念鉴别器与FLaS跨模态融合模块，在16个胸部放射学分类基准及10个定位、分割数据集上实现最优零样本分类性能。
中文摘要 AI 辅助
视觉-语言预训练模型（VLPMs）为开放词汇胸部放射学理解提供了可扩展路径，但仍有两方面未被充分探索：一是从医疗报告中提取的结构化临床语义如何减少对比学习中的批次内噪声，二是如何设计跨模态融合以在不增加复杂度的情况下生成更忠实的空间定位。我们引入AlphaRAD，通过两项贡献解决这些问题。首先，我们构建了由大型语言模型解析医疗报告得到的大规模结构化医疗概念空间用于训练，从而缓解批次内学习噪声，消除对比学习中的启发式配对，使AlphaRAD自然成为通过α校正二元交叉熵训练的医疗概念鉴别器。其次，我们提出FLaS（Factorized Latent Supervision，分解隐层监督），这是一个极其简单却有效的跨模态特征融合模块，它将VLPM表示分解为独立子空间，利用专用对齐监督增强空间定位的表达能力，且不引入额外模型参数。通过大量实证验证，AlphaRAD在各类胸部放射学任务中展现出强大的零样本泛化能力。值得注意的是，它在16个分类基准上取得了最优平均性能，同时在7个定位/短语定位数据集和3个分割数据集上通过不同提升获得了单项最优结果。
英文摘要
Vision-Language Pretrained Models (VLPMs) offer a scalable path to open-vocabulary chest radiology understanding, yet two aspects remain underexplored: how structured clinical semantics extracted from medical reports can reduce in-batch noise during contrastive learning, and how cross-modal fusion can be designed to produce more faithful spatial grounding without added complexity. We introduce AlphaRAD, addressing these opportunities through two contributions. First, we construct a large-scale structured medical concept space from medical reports parsed by a Large Language Model for training, thereby mitigating in-batch learning noise and removing heuristic pair matching in contrastive learning, and thus naturally positioning AlphaRAD as a medical concept discriminator trained via $α$-Corrected Binary Cross-Entropy. Second, we propose FLaS (Factorized Latent Supervision), an extremely simple yet effective cross-modal feature fusion module that factorizes VLPM representations into independent subspaces, using dedicated alignment supervision to enhance the expressiveness of spatial grounding without introducing additional model parameters. Through extensive empirical validation, AlphaRAD shows strong zero-shot generalization across diverse chest radiology tasks. Notably, it establishes state-of-the-art average performance across 16 classification benchmarks, while achieving individual state-of-the-art results via distinct gains on 7 grounding/phrase grounding and 3 segmentation datasets.
发表机构
- Peter Munk Cardiac Centre, University Health Network (UHN)（彼得·芒克心脏中心，大学健康网络（UHN））
- Ted Rogers Centre for Heart Research（泰德·罗杰斯心脏研究中心）
- University of Toronto (U of T)（多伦多大学）
- Vector Institute（向量研究所）
- Toronto General Hospital Research Institute, UHN（多伦多总医院研究所，大学健康网络（UHN））
机构由 AI 辅助整理，请以论文原文为准。
