An electronic nose for insecticides detection in food: the case of alpha-cypermethrin in Swiss chard

Abstract

Due to the disturbingly high quantities of pesticides found in recent years, it has become essential to analyze the food composition intended for consumption. In this study and from a similar angle, we are interested in the detection of alpha-cypermethrin insecticide residues in edible Swiss chard. To this end, we suggest an electronic nose that was constructed using metal oxide gas sensors. Following data collection and pre-processing, two machine learning algorithms—principal component analysis (PCA) and support vector machine (SVM)—were used to analyze the sensor matrix data. The PCA method initially showed that the first three principal components (PCs) may account for more than 96.5% of the sample variation with a clear distinction between known groups corresponding to treated and untreated samples. The identification of untreated Swiss chard from treated one was then accomplished using the SVM method with five folds cross-validation, with a success rate of 92.3%. These results show that our suggestion, which is quick, easy, and affordable, can be utilized as an effective substitute for current methods for identifying Swiss chard that has been treated with the hazardous alpha-cypermethrin.