Multipoint NIR spectrometry and collimated light for predicting the composition of meat samples with high standoff distances

• Collimator aided Multipoint NIR spectroscopy was used to predict the chemical composition of minced beef samples.
• PLSR models were developed for fat, moisture, protein and ash for all the collimator distance–speed combinations.
• Calibration coefficients of determination (Rc2) were obtained in the range of 0.96–0.99.
• Models predicted the independent batch with a good accuracy showing coefficients of determination (Rp2) in the range of 0.92–0.99.

Fat content is one of the most important quality indicators of minced beef products. This study evaluates the efficacy of collimated light and multipoint NIR (Near Infrared) spectroscopy as a rapid, non-destructive and non-contact technique to estimate fat, moisture, protein and ash content of minced beef samples at different measuring distances. A multipoint NIR spectrophotometer, based on a Fabry–Perot interferometer (MultiEye, InnoPharma Labs, Ireland) was used to collect NIR reflectance spectra at three standoff measuring distances (1, 2.5 and 4 cm) from the sample. Measurements were taken in static and rotational motion modes. Prediction models were built using partial least squares regression (PLSR) from the spectral response and proximate analysis. The models for fat content yielded calibration coefficients of determination (Rc2) in the range of 0.96–0.99 with root mean square errors of calibration (RMSEC) in the range of 0.02–4.25 for the three working distances. Good predictions were obtained with root mean square errors of prediction (RMSEP) in the range of 0.03–5.67. Similar results were obtained for the other chemical attributes. Overall results showed good prediction accuracy for all the models. This study demonstrates the ability of multipoint NIR spectroscopy, combined with chemometrics, to predict minced beef composition at increased measuring distances with the aid of collimators.