Congratulations to Xiangyu on the publication of the new work in Engineering Applications of Artificial Intelligence! In this paper, the team proposes CrossToGene, a bidirectional cross-modality interaction framework for predicting spatial gene expression from histopathological images. The method introduces spatial encoding to capture both macro/micro tissue cues, a bidirectional cross-modal interaction module for stronger fusion between image features and gene expression patterns, and contrastive learning to align the visual and molecular spaces. We are also particularly grateful to the reviewers and the Editor for their insightful comments, which have greatly contributed to improving the quality of the manuscript. For more details, please refer to: https://www.sciencedirect.com/science/article/abs/pii/S0952197625037194.

#Computational pathology #Cross-modal learning #Gene expression prediction #Spatial transcriptomics #Tumor microenvironment