Abstract
Fourier transform infrared (FTIR) spectroscopy and attenuated total reflection (ATR) technology were successfully applied for the analysis of total cholesterol (TC) in human serum. An appropriate waveband with stability was selected by the moving window partial least-squares (MWPLS) regression through a rigorous modeling process. The selected waveband was 3065 cm-1 to 945 cm-1, the corresponding PLS factor was 15. The modeling effect M-SEPAve, M-SEPStd, M-RP,Ave, and M-RP,Std were 0.408 mmol/L, 0.068 mmol/L, 0.949, and 0.022, respectively. V-SEP and V-RP, were 0.419 mmol /L and 0.932, respectively. It indicated that the method has good prediction effect and stability, and provided valuable references for designing specialized spectroscopic instruments.
Publisher
Trans Tech Publications, Ltd.