Ning Wang1, Meifang Song2, Haike Gu2, Yiyuan Gao1, Ge Yu3, Fang Lv4, Cuige Shi5, Shangming Wang5, Liwen Sun1, Yang Xiao6, Shucheng Zhang1. 1. Department of Health Quality Center, National Research Institute for Family Planning, Beijing, People's Republic of China. 2. Beijing Radiation Center, Beijing Academy of Science and Technology, Beijing, People's Republic of China. 3. Department of Gynecology, Harbin Medical University Cancer Hospital, Harbin, People's Republic of China. 4. Reproductive Medicine Center, Department of Obstetrics and Gynecology, the Second Affiliated Hospital of Soochow University, Soochow, People's Republic of China. 5. Department of Cell Biology, National Research Institute for Family Planning, Beijing, People's Republic of China. 6. Graduate School of Chinese Academy of Agricultural Sciences, Beijing, People's Republic of China.
Abstract
Objective: The objective of this study is to reduce the dimension of several indicators with a strong correlation when conducting semen quality analysis in a small number of comprehensive variables that could retain most of the information in the original variables. Methods: A total of 1132 subjects were recruited from the Maternal and Child Health Institutions of seven provinces in mainland China. They completed the questionnaire and provided semen samples. Visualization of the correlation between variables was realized by using a function chart and correlation in the PerformanceAnalytics package of the R programming language (version 3.6.3 [2020-02-29]). Factor analysis was conducted using the principal function in the psych package of R. Principal component analysis, combined with varimax rotation, was used in the operation of the model, and two common factors were selected and measured to provide values for the common factor. The score coefficient was estimated using the regression method. Results: The contribution rates of the two common factors to variable X were 43.7% and 33.98%, respectively. When the two common factors were selected, approximately 78% of the information of the original variables could be explained. The correlation coefficients between the first common factor (the quantitative factor) and sperm density, total sperm count, and semen volume were 0.824, 0.984, and 0.544, respectively. The correlation coefficients between the second common factor (the quality factor) and sperm motility and the percentage of forward-moving (progressive spermatozoa) sperm were 0.978 and 0.976, respectively. Conclusion: The correlation between the original variables of a semen quality analysis was strong and suitable for dimensionality reduction by factor analysis. Factor analysis and dimensionality reduction provide a fast and accurate assessment of semen quality. Patients with low fertility or infertility can be identified and provided with corresponding treatments.
Objective: The objective of this study is to reduce the dimension of several indicators with a strong correlation when conducting semen quality analysis in a small number of comprehensive variables that could retain most of the information in the original variables. Methods: A total of 1132 subjects were recruited from the Maternal and Child Health Institutions of seven provinces in mainland China. They completed the questionnaire and provided semen samples. Visualization of the correlation between variables was realized by using a function chart and correlation in the PerformanceAnalytics package of the R programming language (version 3.6.3 [2020-02-29]). Factor analysis was conducted using the principal function in the psych package of R. Principal component analysis, combined with varimax rotation, was used in the operation of the model, and two common factors were selected and measured to provide values for the common factor. The score coefficient was estimated using the regression method. Results: The contribution rates of the two common factors to variable X were 43.7% and 33.98%, respectively. When the two common factors were selected, approximately 78% of the information of the original variables could be explained. The correlation coefficients between the first common factor (the quantitative factor) and sperm density, total sperm count, and semen volume were 0.824, 0.984, and 0.544, respectively. The correlation coefficients between the second common factor (the quality factor) and sperm motility and the percentage of forward-moving (progressive spermatozoa) sperm were 0.978 and 0.976, respectively. Conclusion: The correlation between the original variables of a semen quality analysis was strong and suitable for dimensionality reduction by factor analysis. Factor analysis and dimensionality reduction provide a fast and accurate assessment of semen quality. Patients with low fertility or infertility can be identified and provided with corresponding treatments.
Authors: Joanna Jurewicz; Michał Radwan; Wojciech Sobala; Danuta Ligocka; Paweł Radwan; Michał Bochenek; Wojciech Hanke Journal: Syst Biol Reprod Med Date: 2013-09-30 Impact factor: 3.061
Authors: Feiby L Nassan; Tina K Jensen; Lærke Priskorn; Thorhallur I Halldorsson; Jorge E Chavarro; Niels Jørgensen Journal: JAMA Netw Open Date: 2020-02-05