Use of a portable near-infrared tool for rapid on-site inspection of freezing and hydrogen peroxide treatment of cuttlefish (Sepia officinalis). (February 2022)

• Record Type: Journal Article
• Authors: Currò, Sarah; Fasolato, Luca; Serva, Lorenzo; Boffo, Luciano; Ferlito, Jacopo Carlo; Novelli, Enrico; Balzan, Stefania
• Is Part Of: Food control. Volume 132(2022)
• Publication Date: 2022-02
• Subjects: Cephalopod -- Infrared spectroscopy -- Machine learning -- Food control -- Discrimination; Food -- Quality -- Periodicals; Food -- Analysis -- Periodicals; Food handling -- Periodicals; Food industry and trade -- Quality control -- Periodicals; Aliments -- Industrie et commerce -- Qualité -- Contrôle -- Périodiques; Aliments -- Qualité -- Périodiques; Aliments -- Analyse -- Périodiques; Hygiène alimentaire -- Périodiques; Food -- Analysis; Food handling; Food -- Quality; Periodicals; Electronic journals; 664.07
• DOI: 10.1016/j.foodcont.2021.108524
• Languages: English
• ISSNs: 0956-7135

Abstract

Abstract: Seafood is the product most susceptible to fraudulent activities. In particular, the cephalopod sector is affected mainly by the replacement of fresh by frozen-thawed product either immediately or after bleaching with hydrogen peroxide. Such activities undermine the consumers' confidence in the fish industry and in the effectiveness of the government's food control programme. Currently, the conventional analyses performed to check food authenticity are time-consuming and require skilled personnel, and thus are not applicable to on-site control. The aim of the present study was to develop classification models to discriminate cuttlefish ( Sepia officinalis ) according to physical status and hydrogen peroxide treatment among fresh, frozen-thawed and frozen-thawed hydrogen peroxide-treated samples. A total of 669 cuttlefish spectra were collected, under real conditions, using a portable near-infrared spectrometer (902–1680 nm) operating in reflectance mode. Classification models were developed and then validated using a Support Vector Machine algorithm. Performance of the models was tested through external and hold-out validation according to their accuracy, sensitivity, specificity and Matthew correlation coefficient. The best performance was observed in discriminating between fresh and frozen-thawed product, in which were observed in hold-out validation 0.97 in accuracy and sensitivity, 0.96 in specificity, and 0.92 in Matthew correlation coefficient; whereas, an accuracy of 0.93, sensitivity of 0.89, specificity of 0.94 and Matthew correlation coefficient of 0.83 were found in external validation. However, sample classification according to bleaching treatment showed high performance when fresh samples were excluded. Such achievements confirm the applicability of near-infrared spectroscopy to the on-site inspection of cuttlefish and the control of the fraudulent activity of replacing fresh fish with product mislabelled as such. Highlights: Portable near-infrared tool for a fast on-site control inspection of cuttlefish. Fresh and thawed check method to protect consumers from mislabelling frauds. Discrimination capability of peroxide hydrogen treatment of thawed product. The use of machine learning allows accurate discrimination between groups.