# Shape matters: inferring the motility of confluent cells from static images

## TL;DR

This paper shows how machine learning can predict cell movement in dense layers using only cell shape features from static images.

## Contribution

The study introduces a method to classify cell motility using individual cell shape features rather than collective ones.

## Key findings

- A machine-learning model can accurately predict cell motility using only single-cell shape features.
- Neural networks trained on shape features can classify high- and low-motility cells when motility differences are significant.
- The method works best when the number of highly motile cells is low.

## Abstract

Cell motility in dense cell collectives is pivotal in various diseases like cancer metastasis and asthma. A central aspect in these phenomena is the heterogeneity in cell motility, but identifying the motility of individual cells is challenging. Previous work has established the importance of the average cell shape in predicting cell dynamics. Here, we aim to identify the importance of individual cell shape features, rather than collective features, to distinguish between high-motility and low-motility (or zero-motility) cells in heterogeneous cell layers. Employing the cellular Potts model, we generate simulation snapshots and extract static features as inputs for a simple machine-learning model. Our results show that when cells are either motile or non-motile, this machine-learning model can accurately predict a cell's phenotype using only single-cell shape features. Furthermore, we explore scenarios where both cell types exhibit some degree of motility, characterized by high or low motility. In such cases, our findings indicate that a neural network trained on shape features can accurately classify cell motility, particularly when the number of highly motile cells is low, and high-motility cells are significantly more motile compared to low-motility cells. This work offers potential for physics-inspired predictions of single-cell properties with implications for inferring cell dynamics from static histological images.

We use machine learning to predict single-cell motility in heterogeneous layers using shape features. Our model distinguishes low- from high-motility cells, with potential for physics-inspired predictions from static images of cell dynamics.