Quantitation of hepatitis B virus deoxyribonucleic acid (HBV DNA) is the most accurate method for determining the infection level of hepatitis B. It is a very meaningful means of detecting the different replication states of HBV infection. In this study, HBV DNA was detected and quantified in 499 patients with hepatitis B by confocal Raman spectroscopy combined with multivariate analysis. The effective information related to HBV DNA in spectral data was extracted by different pretreatments according to the vibrational modes of different functional groups. Quantitative analysis of HBV DNA was carried out by using the spectrum in the characteristic range of 807.027–1700.29 cm−1 for spectral peak assignment analysis and using partial least squares to construct a quantitative model. The results show that the regression model is established by combining normalization pretreatment and multivariate analysis methods. The correlation coefficient (Rc) and root mean square error of calibration from the calibration set of the selected feature range are 0.9751 and 0.3881, and the correlation coefficient (Rp) of the verification set and the root mean square error of prediction from the prediction set are 0.9673 and 0.4314. respectively. In clinical medical applications, Raman spectroscopy combined with diversified analysis methods has great research potential for the quantitative detection of HBV DNA in HBV serum.
Read full abstract