Calibration robustness improvements for the early prediction of the pH24 of pork with Raman spectroscopy

The Raman spectroscopy ability for the early prediction of ultimate pH of pork meat is a key feature for French slaughterhouses expectations about meat quality sorting. A previous work has been performed in France at 30 minutes post mortem with a device developed by H. Schmidt (Bayreuth University, Germany) and results were promising for both early prediction of ultimate pH (R²c=0.82) and drip loss (R²c=0.65). This dataset was large enough (n=206), but was lacking DFD samples (pH24 > 6.0; n=4).
The aim of this project is to improve the robustness of Raman prediction models with new samples with high pH24 values. We also tested the ability of the Raman spectrometer to predict pH24 online, with a reduced integration time.