Method development for the analysis of total bacterial count in raw milk using near-infrared spectroscopy

The possibility use of near-infrared (NIR) spectroscopy in the region of 9000-4000 cm(-1) to determine total bacterial count (TBC) in raw milk and the influence of bacterial cultivation on TBC prediction accuracy were studied. Raw milk samples were diluted 10-fold with tryptic soy broth and then cultivated at 32 degrees C for 2 hr. NIR measurements were performed after cultivation times of 30, 60, 90, and 120 min. Partial least squares regression was used to develop calibration models for TBC. The best calibration result was obtained from the 90 min-cultivation samples, with the correlation coefficient (R) of .90 and root mean square error of prediction of .32 log(10) CFU/ml. The analysis of the calibration structure indicated that important information for TBC prediction was from chemical compositional changes and bacterial metabolites. The NIR method combined with bacterial cultivation was satisfactory for the rapid analysis of TBC in raw milk. Practical applicationsThe number of bacteria in raw milk is an important indicator for hygienic quality and milk grading system which consequently reflects a milk price. The proposed method allows satisfied performance for TBC determination in raw milk in 90 min as compared to the 48 hr required using the plate count method. The NIR method developed in this study seems to be an effective tool for a rapid evaluation of bacterial contamination in raw milk. An integration between determination of milk hygienic quality and chemical compositional quality using NIR spectroscopy will facilitate a rapid and multicomponents analysis and become a powerful tool for a practical use in the competitive dairy industry.