Bioimpedance as an alternative tool for subjective, visual scoring of a prevalent ham quality defect

TL;DR

This study explores using bioimpedance as a more reliable alternative to subjective visual scoring for detecting a common pork quality defect.

Contribution

The study demonstrates that bioimpedance correlates strongly with visual scoring of PSE-like defects in pork hams.

Findings

Bioimpedance (Py) showed the strongest correlation with visual scoring (r = −0.46, R2 = 21%).

A multivariate model using Py, pHu, and CIELAB L*a*b* had a prediction error of 0.76, close to the subjective observer error of 0.85.

Combining Py with pH and/or L*a*b* could improve the prediction of PSE-like defects.

Abstract

The detection of meat quality defects can involve both subjective and objective methods. PSE-like meat is linked to a common pork defect and can be caused by rapid post-mortem damage of muscle fibers. This damage can again be linked to various factors, such as a low ultimate pH or a higher slaughter weight. PSE-like defects are characterized by discoloration, structural damage, and excessive moisture loss. However, the lack of suitable instrument-based methods makes the detection of PSE-like defects difficult, and subjective methods typically suffer from poorer reproducibility. The objective of this study was to establish how subjective visual evaluation correlates with electrical impedance spectroscopy and with traditional quality parameters. To do so, visual scoring was performed together with measurements of bioimpedance, color, and pH in two ham muscles (Adductor, Semimembranosus) for 136 animals 24-hours post-mortem. When comparing with visual scoring, Pearson correlation analysis shows the strongest correlation for bioimpedance (Py, r = −0.46, R2 = 21%), followed by pHu (r = 0.44, R2 = 19%). When using all five quality measures, i.e., Py, pHu, and CIELAB L*
a*
b*, the multivariate regression model had a prediction error of 0.76 for the visual scores. This was close to the error describing the subjective bias of visual scoring, more specifically the prediction error between the two observers (0.85). In all, Py showed the strongest correlation among instrument-based quality tests and alone may be used for predicting pork ham structural defects, i.e., as an instrument-based alternative for subjective, visual scoring. However, an instrument that combines Py with pH and/or L*a*b* would improve the prediction of PSE-like quality defects.

Figures