About
Motivation
Cell tracking is a critical task in biological research, enabling scientists to analyze cellular behaviors over time. While deep learning has dramatically improved tracking accuracy, most existing methods are designed for high-frame-rate imaging, where multiple images are captured per second. However, this approach can introduce photobleaching and phototoxicity, potentially affecting biological systems and limiting experimental possibilities.
Low-frame-rate imaging offers a promising alternative, reducing imaging-induced stress and making large-scale cell analysis more feasible. Despite its advantages, automated tracking under these conditions remains a significant challenge.
This repository introduces a new dataset specifically designed for low-frame-rate cell tracking. It features long-term image sequences acquired over one to two days at varying magnifications, complete with ground-truth annotations for cell identification, mitosis, and movement. This dataset enables rigorous evaluation of tracking algorithms in the challenging setting of low-frame-rate data.
Funding
The creation of this website and the companion dataset was supported by the National Institute Of General Medical Sciences of the National Institutes of Health under Award Number R21GM150066. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health.