It is known that specific compounds are produced when foods spoil. For example, commonly recognized spoilage markers include dimethyl sulfi de for chicken and eggs, diacetyl for orange juice, and trimethylamine for fish and milk. As the entire headspace of the spoiled food changes it is possible to detect the spoilage degree by measuring the amount present of these markers. A fast and accurate technique using a mass spectrometry based chemical sensor is examined for the above food products with different spoilage markers’ concentration levels.
Multivariate statistics were used to create models that detect the spoilage markers. Exploratory analysis such as principal component analysis (PCA) and hierarchical cluster analysis (HCA) indicated the viability of the data set for classification models. Soft-independent-modeling-classanalogy (SIMCA) and K Nearest Neighbors (KNN) were used to create two classification models. Regression models were developed using partial least squares (PLS).
Both SIMCA and KNN provided a quick and accurate identification of the above foods with and without spoilage markers. In both cases, testing sets were correctly classified with over 95% accurate prediction rates. PLS models detected spoilage markers’ concentrations at the low to medium ppm levels. Overall, the positive and fast identification of spoilage indicators demonstrates the usefulness of the MS chemical sensor detecting samples with close chemical composition.