Abstract
Multicolor cell spatio-temporal image data have become important to investigateorgan regeneration, malignant outgrowth or immune responses by tracking differentcell types both in vivo and in vitro. Statistical modeling of image data from commonlongitudinal cell experiments poses significant challenges due to the presence of com-plex spatio-temporal interactions between different cell types and difficulties relatedto measurement of single cell trajectories. Current analysis methods oversimplify theproblem to the benefit of computational feasibility, often not providing a full statis-tical treatment of the spatio-temporal effects affecting cell growth. In this paper, wepropose a conditional spatial autoregressive model to describe multivariate count celldata and develop a new parameter estimator within the composite likelihood infer-ence framework. The proposed methodology is computationally tractable and enablesresearchers to estimate a complete statistical model of multicolor cell growth. Theproposed methods are supported by real experimental data where we investigate howinteractions between cancer cells and fibroblasts, which are normally present in thetumor microenvironment, affect their growth.