Fast diagnosis of men's fertility using Raman spectroscopy combined with chemometric methods: An experimental study

Abstract

Background: Idiopathic infertile men suffer from unexplained male infertility; they are infertile despite having a normal semen analysis, a normal history, and physical examination, and when female infertility factor has been ruled out.

Objective: The present study aimed to develop a metabolic fingerprinting methodology using Raman spectroscopy combined with Chemometrics to detect idiopathic infertile men vs. fertile ones by seminal plasma.

Materials and methods: In this experimental study, the seminal plasma of 26 men including 13 fertile and 13 with unexplained infertility who reffered to, Avicenna Infertility Clinic, 2018, Tehran, Iran, have been investigated. The seminal metabolomic fingerprinting was evaluated using Raman spectrometer from 100 to 4250 cm-1. The principal component analysis and discriminate analysis methods were used.

Results: The total of 26 samples were divided into 20 training and 6 test sets. The Principal component analysis score plot of the training set showed that the data were perfectly divided into two sides of the plot, which statistically approves the direct effect of semen metabolome changes on the Raman spectra. A classification model was constructed by linear discriminant analysis using the training set and evaluated by the test group which resulted in completely correct classification. While three of the six test samples appeared in the fertile group, the rest appeared in the infertile as expected.

Conclusion: Metabolic fingerprinting of seminal plasma using Raman spectroscopy combined with chemometric classification methods accurately discriminated between the idiopathic infertile men and the fertile ones and predicted their fertility type.

Keywords: Fertility; Metabolomics.; Raman spectroscopy; Semen analysis.

Figure 1

The Raman spectra (a) before and (b) after autoscaling.

Figure 2

The scores plot of the training set; the red circles indicate the fertile group and the blue idiopathic infertile group (a) before and (b) after autoscaling.

Figure 3

The scores plot of both sets, the training and test; the red stars indicate the fertile group and the blue idiopathic infertile group. The stars are the test set samples.