Forum 03/31/2005

“Reliable Tracking of Large Scale Dense Anti-parallel Particle Motion for Fluorescence Live Cell Imaging”

Ge Yang – Laboratory for Computational Cell Biology, TSRI

This talk is intended to serve two purposes. First, it gives a brief overview from a user’s perspective of advanced particle tracking techniques, including e.g. GNN (global nearest neighbor) and MHT (multiple hypothesis testing). These techniques are highly useful to many fluorescence cellular and molecular imaging applications for automated quantitative tracking of dynamic cellular functions. Second, it introduces the particle tracking technique we developed for FSM (fluorescent speckle microscopy). In particular, it focuses on the tracking of dense antiparallel microtubule flux in meiotic /Xenopus/ egg extract spindles. The basic tracking algorithm integrates motion models at particle, local and global levels. It establishes correspondence between particles based on state similarity and resolves correspondence conflicts using bipartite graph assignment. The tracking algorithm depends critically on computing the global vector field of particle flow, which is carried out iteratively using a new optimal-flow minimum-cost graph algorithm. This technique has been successfully applied to the tracking of speckles in synthetic and real FSM movies and vesicles in Drosophila segmental nerve axon fluorescence movies.