Automated nuclei tracking in C. elegans based on spherical model fitting with multiple target tracking

We present a method for tracking nuclei of C. elegans cells in three dimensional time lapse (4D) data during embryogenesis. It is based on spherical model fitting with multiple target tracking. The algorithm tracks nuclei in a local search space, detects cell division, and searches for newborn sister cells locally. It tracks multiple candidates and postpones the decision to select the newborn sister cell until it becomes bright. This approach results in low error rates. We evaluate the accuracy of the algorithm by comparing to an existing approach on 4D data sets acquired from laser scanning confocal microscopes.