The complex and dynamic nature of cell migration is due to the rapid and continuous remodelling of cytoarchitecture. During this process, spatiotemporal regulation of signalling networks bridges external stimuli to cellular responses. Rho-family GTPases are key components of these signalling networks and most members of this protein family act as molecular switches that cycle between a GDP-loaded form and GTP-loaded form. By applying Raichu FRET probes (with a CFP/YFP fluorescent protein pair) in live endothelial cells, this study develops a novel pixel-clique statistical methodology to quantify the subcellular distribution of Rho-GTPases activity. This new imaging tool is used to observe the Rho GTPase activities during HUVEC cell membrane protrusion. The Matlab statistical toolbox™ was used to implement a linear mixed effect model using the function fitlme[1]. The function ‘getFRET[1]’ performs the analysis with the input images of Raichu-probes expressing cells and single-colour probe expressing cells. In this study, multiple protrusions and retracting membrane segments co-exist in one given cell, which agrees with Yamao and colleagues’ tug-of-war model and Insall’s pseudopod-centred view of cell migration. With this statistical tool, we can begin to examine more in-depth as to how periodic changes in membrane motion is correlated with Rho GTPase activity. The methodology can thus be extended to other (especially non-canonical) Rho GTPase members. An example will be to apply the pixel-clique model to dissect the distinctive and functional overlapping roles of canonical Rac1 and the haematopoietic cell specific Rac2. With the advance in probe design to distinguish highly homologous Rac2 from Rac1 and the pixel-clique model for data interpretation, we can improve our understanding of the diverse roles of Rho GTPases whether in model cell lines, primary cells, stem cells, or in whole small animals.