Data preprocessing and partial least squares regression analysis for reagentless determination of hemoglobin concentrations using conventional and total transmission spectroscopy

Visible-near infrared spectroscopy was successfully used for the determinat ion of total hemoglobin concentration in whole blood. Absorption spectra of whole blood samples, whose hemoglobin concentrations ranged between 6.6 an d 17.2 g/dL, were measured from 500 to 800 nm. Two different types of trans mission were measured: conventional transmission spectroscopy which collect ed primarily collimated radiation transmitted through the sample, and total transmission spectroscopy which used an integrating sphere to collect all scattered light as well, Different preprocessing techniques in conjunction with a partial least squares regression calibration model to predict hemogl obin concentrations were applied to the above two types of transmission, De pending on different preprocessing methods, the standard error of predictio ns ranged from 0.37 to 2.67 g/dL, Mean centering gave the most accurate pre diction in our particular data set. Preprocessing methods designed for comp ensation of the scattering effect produced the worst results contrary to ex pectations. For univariate analysis, better prediction was achieved by tota l transmission measurement than by conventional transmission measurement, N o significant difference was observed for multivariate analysis on the othe r hand, Careful selection of the data preprocessing methods and of the mult ivariate statistical model is required for reagentless determination of hem oglobin concentration in whole blood. (C) 2001 Society oi Photo-Optical Ins trumentation Engineers.