rina-ding/gat-mamba

Combining Graph Neural Network and Mamba to Capture Local and Global Tissue Spatial Relationships in Whole Slide Images # Combining Graph Neural Network and Mamba to Capture Local and Global Tissue Spatial Relationships in Whole Slide Images ### Published journal paper Link In computational pathology, extracting and representing spatial features from gigapixel whole slide images (WSIs) are fundamental tasks, but due to their large size, WSIs are typically segmented into smaller tiles. A critical aspect of analyzing WSIs is how information across tiles is aggregated to predict outcomes such as patient prognosis. We introduce a model that combines a message-passing graph neural network (GNN) with a state space model (Mamba) to capture both local and global spatial relationships among the tiles in WSIs. The model’s effectiveness was demonstrated in predicting progression-free survival among patients with early-stage lung adenocarcinomas (LUAD). We compared the model with other state-of-the-art methods for tile-level information aggregation in WSIs, including statistics-based, multiple instance learning (MIL)-based, GNN-based, and GNN-transformer-based aggregation. Our model achieved the highest c-index (0.70) and has the largest number of parameters among comparison models yet maintained a short inference time. Additional experiments showed the impact of different types of node features and different tile sampling strategies on model performance. Code: github.com. ## Instructions ### Required packages First, create a pytorch docker container using: ``` docker run --shm-size=2g --gpus all -it --rm -v /:/workspace -v /etc/localtime:/etc/localtime:ro pytorch/pytorch:2.2.0-cuda11.8-cudnn8-devel ``` Then install all packages listed here by running the following commands: ``` chmod +x pip_commands.sh ``` ``` ./pip_commands.sh ``` ### Preprocessing If you would like to use the cohorts (NLST or TCGA), magnification level, and tile size used in our paper (either size 512 by 512 at 10x (1 mpp) or 1024 by 1024 at 20x (0.5 mpp) so that the total area covered by a …