Microtubules are the most dynamic filaments in a cell’s cytoskeleton, where they form networks that span the cell and have been implicated in cell migration, mitosis, and intracellular transport. Research into microtubule dynamics relies heavily on various microscopy and low throughput image processing and analysis that have been limited to the investigation of single microtubules or stationary cells. We have written an image analysis program in MATLAB capable of detecting fluorescent microtubules in time-lapse confocal microscopy images of motile cells with minimal manual input. The semiautomatic program constructs a worm-like chain from an initial guess and optimizes the chain placement relative to the image intensity profile of a localized region. In subsequent frames, gradient vector flow guides the chain toward the microtubule despite conformational changes. Efficient extraction of dynamic microtubule morphologies serves as a key first step to investigating the forces experienced by these cytoskeletal elements in the cytoplasm of crawling cells.
Author entry (protected)