An algorithm for estimation of chromosome motion in Four-Dimensional microscopic images
Analysis of the three-dimensional organization of chromosomes within the nucleus has revealed a number of characteristic structural features. Yet imaging of living nuclei indicate that chromosomes undergo considerable random motion. Maintenance of nuclear organization in the face of such motion is thought to involve the attachment of chromosomes to the nuclear envelope or matrix. Yet while such attachments have been proposed to play a variety of functional roles as well as maintain nuclear organization, direct evidence for the existence of these mechanical interactions in vivo has been lacking. One way to demonstrate such attachment directly would be to estimate the motion of chromosomes and attempt thereby to demonstrate the presence of fixed points, which would indicate attachment of chromatin to some fixed superstructure.We have previously presented a motion estimation algorithm that is designed for tracking the motion of nonrigid and featureless objects such as chromosomes. This algorithm starts with a structural representation for the set of chromosomes at each time point, and then finds a correspondence between elements of the representations at successive time points.