Controlling endogenous angiogenesis, or the formation of new capillaries, is a potential therapeutic strategy for numerous diseases including cancer, stroke and cardiovascular disease (1–4). These efforts have been met with mixed success in the clinic, partly due to an inadequate understanding, and thus control, of the mechanisms that influence the endothelial cells that form capillaries (1, 3). In order to control angiogenesis in an effort to improve treatment responses, quantitative information about endothelial cell behavior must be used to build accurate models of vascular network formation.
In this paper, we introduce a method to identify and classify endothelial cell responses to angiogenic stimuli through sophisticated image analysis. The presented automated image processing tools and classification framework allow for rapid quantitative investigations of cellular images. Results of our analysis demonstrate that endothelial cells can be grouped into distinct morphological phenotypes as a function of their responses to combinations of angiogenic growth factor stimuli. Information on phenotypic behavior and responses will be applied towards predicting and guiding cell behavior for therapeutic design.