MULTIPLE HYPOTHESIS TRACKING WITH INTEGRATED CELL DIVISION DETECTION

Automatic tracking of proliferating cells in microscopy images is important to elucidate biological processes. We have developed a new probabilistic approach for cell tracking which is based on Multiple Hypothesis Tracking and integrates cell division detection. Our method uses information from multiple frames and formulates data association with cell division detection as graph-theoretical maximum weighted independent set problem. We evaluated our approach using synthetic data as well as data from the Cell Tracking Challenge. It turned out that our approach generally improves the results for cell tracking and cell division detection compared to previous methods.