Information from the six systems was used independently and some instruments, Visible-NIR, pH, colorimeter and electrical impedance, were used in combination to predict pork tenderness. pH, colorimeter, electrical impedance and Visible-NIR instruments were also used to predict pork loin chop lipid and pH. The intent was that if a system or systems could be used to predict tenderness, could the system(s) also be used to predict pH and marbling.
The Tenera Technology High Resolution Imaging System was relatively new and only 376 pork chop images were analyzed due to mechanical design flaws. This technology is being redesigned to capture features at an ultra-structural level and has potential to enhance pork tenderness prediction. Further development is being conducted at Colorado State University.
The Visible-NIR system effectively predicted pork tenderness alone using partial least squares regression models. When applied to a calibration data set, Visible-NIR correctly classified 97% of tender, 93% of intermediate, and 92% of tough pork chops based on Warner-Bratzler shear force measurements of tenderness. When Visible-NIR was used in combination with pH and/or colorimeter data, predictions were improved. However, the improvements in prediction by adding additional instruments would most likely not offset additional labor costs, instrument costs and issues with using multiple instruments at processing plant line speeds. The Visible-NIR system could also be used to assess loin pH and lipid (marbling) content with similar accuracy as pork tenderness. Stress-Strain Imaging shows promise and could potentially be combined with existing commercial ultrasound instrumentation for development of prediction models.