Nafisi Vahid Reza, Moradi Mohammad Hasan, Nasr-Esfahani Mohammad Hosain
Bioelectric Department, Faculty of Biomedical Engineering, Amirkabir University of Technology, Tehran, Iran.
Physiol Meas. 2005 Oct;26(5):639-51. doi: 10.1088/0967-3334/26/5/006. Epub 2005 Jun 13.
The conventional assessment of human semen, especially sperm movement characteristics, is a highly subjective assessment, with considerable intra- and inter-technician variability. Computer-assisted sperm analysis (CASA) systems provide a rapid and automated assessment of the parameters of sperm motion, together with improved standardization and quality control. However, it should be noted that the measurement of the sperm head motion by CASA is sensitive to the technique of experimentation. While conventional CASA systems use digital microscopes with phase-contrast accessories that make the sperm's head appear brighter and sharper than the other parts, in this research, a regular light microscope was used with a digital camera directly attached to its eyepiece. One of the drawbacks of this method is that the images lack proper contrast and sharpness. To remedy this, we have proposed an algorithm for sperm tracking that is insensitive to image acquisition conditions. This tracking algorithm was used after the background and extra particles were successfully removed through a two-step enhancement algorithm. Additionally, in this research, a template matching method was used for finding the sperm's path. Upon examination, it was proven that our tracking algorithm worked well with different image acquisition conditions. This paper explains how this method reduces error probability in finding and tracking sperm in various frames.
对人类精液的传统评估，尤其是精子运动特征的评估，是一种高度主观的评估，技术人员内部和之间存在相当大的差异。计算机辅助精子分析（CASA）系统可对精子运动参数进行快速、自动的评估，同时提高了标准化程度和质量控制。然而，应该注意的是，CASA对精子头部运动的测量对实验技术很敏感。传统的CASA系统使用带有相差附件的数字显微镜，使得精子头部比其他部分看起来更亮、更清晰，而在本研究中，使用的是直接在目镜上连接数码相机的普通光学显微镜。这种方法的一个缺点是图像缺乏适当的对比度和清晰度。为了弥补这一点，我们提出了一种对图像采集条件不敏感的精子跟踪算法。在通过两步增强算法成功去除背景和额外颗粒后，使用了这种跟踪算法。此外，在本研究中，使用模板匹配方法来寻找精子的路径。经检验，证明我们的跟踪算法在不同的图像采集条件下都能很好地工作。本文解释了该方法如何降低在不同帧中查找和跟踪精子时的错误概率。
