AI 中文总结
本研究提出C²A分类头，通过耦合空间证据与临床先验，在CheXpert数据集上实现0.895的宏观平均AUROC，提升了高度共现模糊类别的分类性能。
AI 中文摘要
胸部病理很少单独出现，但标准多标签分类器依赖共享的全局描述符，会丢弃病灶的位置信息及共现方式。我们提出C²A（共现感知类注意力），一种明确耦合空间证据与临床先验的分类头。首先，C²A将池化转化为对学习到的每类空间注意力图的期望，生成每种疾病的局部描述符；其次，它通过从经验标签共现初始化的可学习图来耦合这些描述符，单步残差消息传递在相关病灶间共享证据，该过程是对恒等映射的有界扰动，共现通过显式双线性交互项进入每个logit。在CheXpert数据集上，C²A取得了0.895的宏观平均AUROC，优于先进的上下文门控基线。关键的是，性能提升集中在空间证据模糊的高度共现类别上（使肺不张的AUROC较GCG基线提升1.5），证明了该先验的正则化效果，且仅引入一个线性投影和C×C边矩阵的可忽略开销。
英文摘要
Thoracic pathologies rarely occur in isolation, yet standard multi-label classifiers rely on shared global descriptors, discarding \emph{where} findings lie and \emph{how} they co-occur. We propose \textbf{C$\mathbf{^2}$A} (Co-occurrence Aware Class Attention), a classification head that explicitly couples spatial evidence with clinical priors. First, C$^2$A casts pooling as an expectation over learned per-class spatial attention maps, yielding localized descriptors for each disease. Second, it couples these descriptors via a learnable graph warm-started from empirical label co-occurrence. A single residual message-passing step shares evidence among related findings, proving to be a bounded perturbation of the identity where co-occurrence enters each logit through an explicit bilinear interaction. On CheXpert, C$^2$A achieves a superior $0.895$ macro-mean AUROC, outperforming advanced context-gating baselines. Crucially, gains concentrate on highly co-occurrent classes with ambiguous spatial evidence (rescuing Atelectasis by $+1.5$ over GCG), demonstrating the prior's regularizing effect with a negligible overhead of one linear projection and a $C\!\times\!C$ edge matrix.
CommentsAccepted at 2026 IEEE International Workshop on Machine Learning for Signal Processing
