Image Processing Methods for Automated Assessment of Sperm DNA Integrity

Abstract

Infertility is a rising concern across the world and it is estimated to affect approximately 15% of the couples. Although there are many factors responsible for infertility, male infertility constitutes 50% of the cases. Male fertility is largely dependent on sperm quality. Sperm is a specialized cell and the fertilization potential of a sperm cell relies on the integrity of sperm DNA, apart from other factors. A set of seminal analyzes is done in a traditional way to determine the quality of sperm cells, but these have limited capability of detecting DNA damages. The aim of this study is to develop a novel image processing technique for automated, cost–effective, and rapid assessment of sperm cell DNA damage for addressing infertility issues. The microscopic images of sperm cells were generated using the Giemsa staining procedure. The k-means clustering method was applied on the images to segment and separate the core and halo parts of the sperm cell. Using centroid-based measures, the difference in diameters between the core and halo parts were calculated. Based on the range of diameter differences, assessment was made on the number of sperm cells with small halo, medium halo, big halo, and no halo. The percentage of degraded cells was represented as the fraction of cells having no or small halos as compared to the ones with big halos. A set of ten, real-time microscopic images of semen samples were considered in this study. The results are suggestive of the potential of the proposed method for rapid identification of degraded sperm cells.