Image Analysis of the Motility of Spermatozoa

Martin Proença

Minor project in Biomedical Technology Project

Microengineering section, EPFL

January 2010

Abstract

During this Minor project in Biomedical Technology, an ImageJ plugin allowing the analysis of the motility of spermatozoa has been developed. It has been designed for microscope images acquired with a 20x objective. The goal is to provide a tool to biologists to help them in computing a fertility index based on the global behaviors observed among the sperms (progressively motile, non-progressively motile, non-motile, etc.). The approach implements a case-adapted version of a fast Multiple Hypothesis Tracking technique. It aims not only at coping with false and missed detections, but also at performing a faster tracking by a huge parallel computation. The user can freely adjust some parameters such as the sizes of the filter that detects the heads of the sperms, the depth of the algorithm (i.e. the number of future frames taken into account for the association problem) and the search radius, which depends on the type of sperms one is dealing with. The final output of this plugin is a results table containing the coordinates of the trajectories at each frame. The analysis of these trajectories can then lead to a fertility index.