Image segmentation techniques for stem cell tracking

To grow stem cells in vitro is an important task in regenerative medicine. Cell motility that can be derived by cell tracking is a useful index to evaluate the viability and stemness of stem cells. The precision of cell tracking is highly dependent on correct detection of the cell centroids, which are usually determined by cell segmentation process. In this study, a cell segmentation method combining double-thresholding and disk-based reconstruction is proposed to solve the problems arising from shape deformation and uneven illumination during cell culture. The qualitative and quantitative comparisons show that the proposed segmentation method increases accuracy in detection the cell centroid.