Unusual structures inherent in point pattern data predict colon cancer patient survival

TL;DR

This study introduces a novel point process approach to analyze cell distribution patterns in cancer tissue images, enabling prediction of colon cancer patient survival based solely on spatial cell arrangements.

Contribution

It develops and applies a new point process methodology, including a void process, to link tissue cell spatial patterns with patient prognosis, advancing quantitative tissue analysis.

Findings

Cell spatial patterns can predict patient survival.

New point process models effectively fit tissue data.

Spatial analysis improves prognostic accuracy.

Abstract

Cancer patient diagnosis and prognosis is informed by assessment of morphological properties observed in patient tissue. Pathologists normally carry out this assessment, yet advances in computational image analysis provide opportunities for quantitative assessment of tissue. A key aspect of that quantitative assessment is the development of algorithms able to link image data to patient survival. Here, we develop a point process methodology able to describe patterns in cell distribution within cancerous tissue samples. In particular, we consider the Palm intensities of two Neyman Scott point processes, and a void process developed herein to reflect the spatial patterning of the cells. An approximate-likelihood technique is taken in order to fit point process models to patient data and the predictive performance of each model is determined. We demonstrate that based solely on the spatial arrangement of cells we are able to predict patient survival.