Detecting spoilage of tomatoes by ultrasound spectroscopy
University of New Brunswick
Food spoilage caused by bacterial or fungal contamination can lead to cases of food poisoning and even death. Current techniques for spoilage detection cannot routinely measure packaged products. We present how ultrasound spectroscopy can be used to classify spoiled and non-spoiled samples of tomato juice. Spoilage was induced with a bacterium (Lactobacillus rhamnosus) and mold (Geotrichum candidum) commonly found in spoiled tomatoes and samples analyzed using ultrasound spectroscopy. Frequency profiles of the inoculated samples showed changes with spoilage when compared to control samples. Cross-validated multivariate models separated fresh and spoiled samples with 100% accuracy. A global model was created using the combined inoculation data resulting in a sensitivity and specificity each of 94% with an independent validation set resulting in 100% accuracy of separation. Overall, ultrasound spectroscopy shows promise in the detection of food spoilage. The potential for measuring food samples non-invasively could benefit both the food industry and consumers.