Non-invasive quantitative analysis of human blood components in renal three items using spectral compensation method based on spectral data and component content correlation. (December 2022)

• Record Type: Journal Article
• Authors: Wang, Kang; Li, Gang; Wang, Dan; Lin, Ling
• Is Part Of: Computer methods and programs in biomedicine. Volume 227(2022)
• Publication Date: 2022-12
• Subjects: Noninvasive -- Spectral compensation method -- Dynamic spectrum and "M+N" theory -- Urea nitrogen -- Creatinine -- Uric acid; Medicine -- Computer programs -- Periodicals; Biology -- Computer programs -- Periodicals; Computers -- Periodicals; Medicine -- Periodicals; Médecine -- Logiciels -- Périodiques; Biologie -- Logiciels -- Périodiques; Biology -- Computer programs; Medicine -- Computer programs; Periodicals; Electronic journals; 610.28
• DOI: 10.1016/j.cmpb.2022.107210
• Languages: English
• ISSNs: 0169-2607

Abstract

Highlights: A complete non-invasive quantitative analysis system of human blood components. This paper proposes a spectral compensation method. Based on "M+N" theory, dynamic spectrum, mathematical statistics, and data analysis. This method is a universal method for spectral pretreatment in spectral analysis. Abstract: Background and objective: The non-invasive measurement of the blood component contents in humans using spectroscopic methods has important clinical application value. In the absorption spectrum of human blood, for a target component with a small content, the absorption spectrum line is low, the absorption spectrum line of the target component changes weakly at different wavelengths, and the absorption spectrum lines of the target and non-target components overlap. The above reasons cause the spectral line difference between the target and non-target components to be small. However, non-target components also have physical properties such as scattering, which will reduce the contribution of the target component in the absorption spectrum of blood, that is, the spectral line difference between the target and non-target components becomes smaller, thus making the accuracy of the target component model lower. Methods: To increase the accuracy of modeling analysis, according to the correlation between spectral data and the content of the target component, this paper proposes a spectral compensation method. The method compensates for the influence caused by the non-target components on the target components through the mathematical relations between the data of the spectrum and the contents of the target components. Thus, a high-precision, non-invasive quantitative analysis system of human blood components is obtained. Taking the analysis of the three components of urea nitrogen, creatinine, and uric acid as examples, the experiment used spectral data before and after spectral compensation and partial least squares (PLS) methods to build two models and used them to predict the content of the target component. Results: When the models of different components established by the spectral compensation method respectively predict the all samples' contents of urea, creatinine, and uric acid, the three correlation coefficients can reach more than 0.9700. Conclusion: It can be obtained from the experimental results that the spectral compensation method can greatly increase the performance of each component model. Spectral compensation is a new and comprehensive method for improving the quality of spectral data. Therefore, this paper obtains a non-invasive quantitative analysis system of human blood components with higher precision, and can quantitatively analyze blood urea nitrogen, creatinine, and uric acid in human blood with high precision, which has great practical application value in clinical practice.