It is a critical task to evalaute HER2 expression level accurately for breast cancer evaluation and targeted treatment therapy selection. However, the standard multi-step Immunohistochemistry (IHC) staining is resource-intensive, expensive, and time-consuming, which is also often unavailable in many areas. Consequently, predicting HER2 levels directly from H&E slides has emerged as a potential alternative solution. It has been shown to be effective to use virtual IHC images from H&E images for automatic HER2 scoring. However, the pixel-level virtual staining methods are computationally expensive and prone to reconstruction artifacts that can propagate diagnostic errors. To address these limitations, we propose the Latent-Guided Dual-Stream Network (LGD-Net), a novel framework that employes cross-modal feature hallucination instead of explicit pixel-level image generation. LGD-Net learns to map morphological H&E features directly to the molecular latent space, guided by a teacher IHC encoder during training. To ensure the hallucinated features capture clinically relevant phenotypes, we explicitly regularize the model training with task-specific domain knowledge, specifically nuclei distribution and membrane staining intensity, via lightweight auxiliary regularization tasks. Extensive experiments on the public BCI dataset demonstrate that LGD-Net achieves state-of-the-art performance, significantly outperforming baseline methods while enabling efficient inference using single-modality H&E inputs.

翻译：准确评估HER2表达水平对于乳腺癌评估和靶向治疗方案选择至关重要。然而，标准的多步骤免疫组织化学染色方法资源消耗大、成本高且耗时，在许多地区也难以普及。因此，直接从H&E切片预测HER2水平已成为一种潜在的替代方案。已有研究表明，利用H&E图像生成虚拟IHC图像进行自动化HER2评分是有效的。但像素级虚拟染色方法计算成本高昂，且易产生重建伪影，可能导致诊断误差传播。为克服这些限制，我们提出潜在引导双流网络，该新颖框架采用跨模态特征幻觉而非显式的像素级图像生成。LGD-Net在训练期间通过教师IHC编码器引导，学习将形态学H&E特征直接映射到分子潜在空间。为确保幻觉特征能捕捉临床相关表型，我们通过轻量级辅助正则化任务，显式地利用任务特定领域知识（特别是细胞核分布与膜染色强度）对模型训练进行正则化。在公开BCI数据集上的大量实验表明，LGD-Net实现了最先进的性能，在仅使用单模态H&E输入进行高效推理的同时，显著优于基线方法。