Study on Sperm DNA Integrity Detection Using an Image Processing Approach

Abstract

Human reproductive issues are on the rise across the globe due to change in lifestyle, food habits and deteriorating environmental conditions. Hence, detection and treatment of infertility is gaining impetus. Quality of the sperm plays a crucial role in male infertility. Manual detection of sperm quality is prone to pitfalls. Computer aided sperm analysis is economically distant in developing and under developed countries. Hence, this research is an effort towards developing simple image processing techniques to detect the integrity of sperm DNA. Microscopic images of semen samples from subjects with and without fertility issues were considered. Two image processing techniques, namely, threshold and extended maximum transform and K-means clustering were employed to detect the core and halo parts of the sperm DNA. Based on the difference in the diameters of these parts, the sperm halos were categorized and degraded cells were identified. The sperm cell categorization with K means algorithm provides 96.6% accuracy with the limited number of samples. The results of the study suggest that the proposed method would help in developing simple and economical means of sperm integrity testing.