Skip to content

Latest commit

 

History

History
13 lines (5 loc) · 748 Bytes

File metadata and controls

13 lines (5 loc) · 748 Bytes

histopathology_multi-omics

A multi-modal model that integrates histopathological image data and gene expression profiles to classify breast cancer subtypes (LumA, LumB, Basal, Her2). The model leverages a Graph Mamba encoder for omics data and a mamba-based image encoder built on a pretrained ConvNeXt backbone from timm.

image gex_heatmap gex_boxplot_FGA