Histology analysis of the tumor micro-environment integrated with genomic
assays is the gold standard for most cancers in modern medicine. This paper
proposes a gene-induced multimodal pre-training (GiMP) framewo
整合多模态存活方法、千亿像素组织学全幻灯片图像和转录组学,通过凝聚形态学原型和生物途径原型对 WSI 和转录组简化,再利用聚合网络(Transformer 或最优运输交叉对齐)处理得到的多模态令牌,能在较小且固定的令牌数量下实现高性能预测肿瘤患者预后、分层,同时实现较少的计算量并具备新的可解释性分析。