Graph Based Neural Networks for Interrogating Multiplexed Tissue Samples

Anonymous

Keywords: Multiplexed Immunofluorescence, Graph Neural Networks, Digital Pathology, Colorectal Cancer

Abstract: Multiplexed immunofluorescence provides an unprecedented opportunity for studying specific cell-to-cell and cell-microenvironment interactions. Noise, imaging artifacts, and the variation in protein expression make this a particularly challenging problem. We employ graph neural networks to combine features obtained from tissue morphology with measurements of protein expression to identify communities of cells related to tumour stage. Our framework presents a new approach to analysing and processing these complex multi-dimensional datasets.