Automatic approach for analyzing the endosome movement in fluorescence video microscopy

Endosomes are structures located in a biological cell and their movement can be interesting for studying the effect of pharmaceutical substances on cell processes. The movement of endosomes can be analyzed by using fluorescence microscopy and by producing videos of few seconds length. Movement analysis done by a human specialist is usually unappealing because it is both time consuming and difficult, also the results depend always on the person doing the analysis. By automating the movement analysis we can ease and speed up the process, and the results would also be easily replicable. The paper presents a new tracking approach for endosome movement based on the double exponential smoothing. When compared to Kalman filter, a commonly used movement predictor, the proposed method proved to be simpler to implement, with lower running time and with good results. Also, the paper shows ways to express the movement of the tracked endosomes.