Evaluation of bacterial population on chicken meats using a briefcase electronic nose

A novel portable electronic nose (E-nose) based on eight metal oxide sensors was used for evaluation of chicken meat freshness and bacterial population on chicken meat stored at 4.0 °C and 30.0 °C for up to 5 days. Aerobic plate counts were employed for the total count of bacterial population in term of typical biological analysis. Gas chromatography/mass spectrometry (GC/MS) analysis of fresh and spoilage chicken meats was presented. Principal component analysis (PCA) was used for pattern recognition and classification. A model for bacterial population evaluation was built by using a back propagation neural network (BPNN) based on sensor responses from the E-nose. The PCA results clearly showed the classification of chicken meat freshness corresponding to different storage days and temperatures. The E-nose with a constructed BPNN prediction model exhibited good evaluation of bacterial population on chicken with high correlation coefficient (R2 = 0.94) and mean square error of 0.016. The results suggested that the developed E-nose system can be used as a rapid and alternative way for evaluation of bacterial population on meats and offers several advantages including fast, portable, low cost, and non-destructive measurement with high relative accuracy.