Automated multiple cell tracking method for non-invasive analysis of cells using dielectrophoresis.

For the non-contact manipulation of cells using dielectrophoresis, it is necessary to accurately measure the DEP-induced trajectories of numerous cells at the single-cell level. Accurate cell tracking is essential for image-based characterization of cellular DEP mobility. In this paper, we present a tracking algorithm based on the localization information of Michigan Cancer Foundation-7 (MCF -7) cells detected in every frame. To evaluate the performance of the multi-cell tracking algorithm, we evaluated the tracking accuracy of cells by the DEP force oscillating between positive and negative values. The evaluation index was evaluated using the multi-object tracking accuracy (MOTA), which is used for general performance evaluation of multi-object tracking systems. Cell tracking ability was over 90% based on cells detected in the first frame.