Diagnosis of spoilage in tomato paste by Bacillus subtilis bacteria, Penicillium and Aspergillus fungies using electronic nose
The goal of this research was to investigate the performance and ability of the electronic nose as a new, fast and low-cost tool in detecting spoilage caused by Bacillus subtilis bacteria and Penicillium and Aspergillus fungi in tomato paste, as well as to investigate the changes in some of the physicochemical properties of tomato paste.
In order to perform tomato paste spoilage tests, two strains of fungies, named Penicillium and Aspergillus, and one strain of bacteria named Bacillus subtilis were used. Fungi and bacteria were cultured in solid agar medium. After the cultivation,bacteria and fungies, they were placed in an incubator for 72 hours so that the microorganisms grow in the culture medium. After 72 hours, the microorganisms were removed from the incubator and a suspension with a concentration of 107 cfu/ml was prepared from them and was added to the tomato paste samples free of microorganisms. The incubation period for the growth of microorganisms in tomato paste was carried out as follows.
For samples containing bacteria, 48 hours
For samples containing fungi, one week
Sampling were performed after the incubation period for bacteria-infected samples at 4-hour intervals for 24 hours and for fungies-infected samples at daily intervals for one week. Counting the number of bacteria and fungies in the initial and final sampling was done by the hemocytometer method. And finally, the data were analyzed using QDA, ANN, SVR .
The QDA method has been able to classify the paste samples infected with bacteria based on different sampling times with 100% accuracy. Also, this method was able to classify samples infected with fungies based on different sampling days and type of fungi with 98.57% and 77.86% accuracy, respectively. In general, it can be concluded that the electronic nose with the help of QDA method was able to detect the growth of fungies on different days of sampling, while the electronic nose with the help of QDA method could not perform very well in detecting the type of fungi. The artificial ANN network has been able to classify the paste sample infected with bacteria based on the sampling time with an accuracy of 86.7%. Also, the artificial ANN network was able to classify with 76.4% accuracy the paste sample infected with fungi based on the sampling time. Also, the ANN was able to classify with 90% accuracy the sample of paste infected with fungi based on the type of fungi. The results obtained by ANN for predicting parameters of pH, Brix, acidity and sediment weight percentage of the samples infected with bacteria showed that the highest and lowest value of R2 belonged to the sediment weight percentage and acidity parameters, respectively. Also the highest and lowest value R2 obtained from the SVR method related to the pH and sediment weight percentage parameters. Also, in the samples infected with fungus, the highest value of R2 obtained by ANN and SVR methods belonged to the acidity parameter and the lowest related to the sediment weight percentage parameter.
In this research, the use of an electronic nose system based on ten metal oxide semiconductor (MOS) sensors to detect the spoilage caused by Bacillus subtilis bacteria and Aspergillus and Penicillium fungies in tomato paste was investigated. In order to classify the samples, QDA and ANN methods were used. QDA method with 100% accuracy and ANN with 87.6% accuracy classified the tomato paste sample infected with bacteria based on 6 different sampling times. The accuracy of classifying samples infected with fungies based on the type of fungi by ANN was 90%. According to this research, it can be concluded that the electronic nose is a suitable tool for detecting spoilage caused by Bacillus subtilis bacteria and Aspergillus and Penicillium fungies in tomato paste and it can detect spoilage in tomato paste with less time and cost.