A Hand-Held Platform for Boar Sperm Viability Diagnosis Based on Smartphone

Abstract

The swine fever virus seriously affects pork production, and to improve pork production, pig breeding efficiency needs to be improved, and the detection of boar sperm activity is an important part of the pig breeding process. Traditional laboratory testing methods rely on bulky testing equipment, such as phase-contrast microscopes, high-speed cameras, and computers, which limit the testing scenarios. To solve the above problems, in this paper, a microfluidic chip was designed to simulate sperm in the oviduct with a channel thickness of 20 um, which can only accommodate sperm for two-dimensional movement. A miniature microscope system which can be used in combination with a smartphone is designed that is only the size of the palm of the hand and has a magnification of about 38 times. An intelligent diagnostic app was developed using Java language, which can automatically identify and track boar sperm with a recognition rate of 96.08% and an average tracking rate of 86%. The results show that the proposed smartphone-based hand-held platform can effectively replace the traditional microscope compound computer to diagnose sperm activity. In contrast, the platform is smaller, easier to use and is not limited by the usage scenarios.

Sperm YOLOv8E-TrackEVD: A Novel Approach for Sperm Detection and Tracking

“…Their research achieved a 91.77% accuracy rate in sperm head detection on the VISEM sperm sample video dataset, demonstrating a high correlation (Pearson’s r = 0.969) with laboratory analysis methods. Zheng et al [ 20 ] used Gaussian mixture models to identify moving sperm in videos, Kalman filtering to predict their future positions, and the Hungarian algorithm to match predicted positions with actual detected ones, ensuring accurate tracking of each sperm’s trajectory. These methods may face computational challenges when dealing with a large number of sperm and depend on the quality of images and frame rates, often underperforming with low-quality video data.…”