Histographs: Graphs in Histopathology
This work addresses the challenge of explicitly capturing spatial cell arrangements in histopathology for cancer detection, offering a domain-specific improvement over CNNs.
The authors tackled the problem of classifying cancers from histopathology images by modeling tissue sections as multi-attributed spatial graphs of cells, using graph convolutional networks (GCNs). They achieved classification accuracy competitive with Inception-v3 CNNs on breast cancer tasks, specifically on the BACH dataset for cancerous vs. non-cancerous and in situ vs. invasive classifications.
Spatial arrangement of cells of various types, such as tumor infiltrating lymphocytes and the advancing edge of a tumor, are important features for detecting and characterizing cancers. However, convolutional neural networks (CNNs) do not explicitly extract intricate features of the spatial arrangements of the cells from histopathology images. In this work, we propose to classify cancers using graph convolutional networks (GCNs) by modeling a tissue section as a multi-attributed spatial graph of its constituent cells. Cells are detected using their nuclei in H&E stained tissue image, and each cell's appearance is captured as a multi-attributed high-dimensional vertex feature. The spatial relations between neighboring cells are captured as edge features based on their distances in a graph. We demonstrate the utility of this approach by obtaining classification accuracy that is competitive with CNNs, specifically, Inception-v3, on two tasks-cancerous versus non-cancerous and in situ versus invasive-on the BACH breast cancer dataset.