Somatic cell count determination in cow's milk by near-infrared spectroscopy: a new diagnostic tool

The potential of near-infrared spectroscopy (NIR) in the region from 1,100 to 2,500 nm to measure somatic cell count (SCC) content of cow's milk was investigated. A total of 196 milk samples from seven Holstein cows were collected for 28, consecutive days, starting from 7th d after calving, and analyzed for fat, protein, lactose, and SCC. Three of the cows were healthy, and the remainder had periods of mastitis during the experiment. Near-infrared transflectance milk spectra were obtained using an InfraAlyzer 500 spectrophotometer. The calibration for logSCC was performed using partial least square (PLS) regression and different spectral data pretreatment. The best accuracy of determination was found for an equation that was obtained using smoothed absorbance data and 10 PLS factors. The standard error of calibration was 0.361, the calibration coefficient of multiple correlation was 0.868, the standard error of prediction for independent validation set of samples was 0.382, the correlation coefficient was 0.854, and the coefficient of variation was 7.63%. The accuracy of logSCC determination by NIR spectroscopy would allow health screening of cows and differentiation between healthy and mastitic milk samples. It has been found that SCC determination by NIR milk spectra is based on the related changes in milk composition. The most significant factors that simultaneously influenced milk spectra with the elevation of SCC were alteration of milk proteins and changes in ionic concentration of milk.