HisCMCL: Cross-Modal Contrastive Learning for Spatial Expression Prediction

HisCMCL is a multimodal framework designed to infer spatial transcriptomics (ST) from pathology images by combining multi-scale features with cross-attention and contrastive learning.