Detection of paralytic shellfish toxins by near‐infrared spectroscopy based on a near‐Bayesian SVM classifier with unequal misclassification costs

Abstract

BackgroundParalytic shellfish poisoning caused by human consumption of shellfish fed on toxic algae is a public health hazard. It is imperative to implement shellfish monitoring programs to minimize the possibility of shellfish contaminated by paralytic shellfish toxins (PST) reaching the marketplace.ResultsA rapid detection method for PST in mussels using near‐infrared spectroscopy (NIRS) technology has been proposed. The spectral data in the wavelength range of 950‐1700 nm for PST‐contaminated and non‐contaminated mussel samples have been used for building the detection model. To solve the imbalanced classification problem that the quantity of non‐contaminated mussels is far less than that of PST‐contaminated mussels in practical application, near‐bayesian support vector machines (NBSVM) with unequal misclassification costs (u‐NBSVM) have been applied. The u‐NBSVM model performs competitively on imbalanced datasets by combining unequal misclassification costs and decision boundary shifts. The detection performance of the u‐NBSVM does not show worse with the number of PST samples decreasing by adjusting the misclassification costs. When the number of PST samples is 20, the G‐mean and accuracy can reach 0.9898 and 0.9944, respectively.ConclusionCompared with the traditional SVM and the NBSVM, the u‐NBSVM model achieves better detection performance. The results of this study validate that the NIRS technology combined with the u‐NBSVM model can be used for rapid and non‐destructive PST detection in mussels.This article is protected by copyright. All rights reserved.