Assessment of microplastic contamination in shrimp utilizing multispectral imaging, fluorescence, and infrared spectroscopy

This research explores the potential of three spectroscopic techniques: Mid-Infrared (MIR) spectroscopy, fluorescence spectroscopy, and multispectral imaging (MSI) in predicting and categorizing microplastic (MP) contamination levels in minced shrimp. The shrimp samples were mixed with varying concentrations of polystyrene (PS), polypropylene (PP), polyethylene (PE), and polyethylene terephthalate (PET) microplastics at 11 different levels, ranging from 0% to 10% w/w. These levels included 0%, 0.04%, 0.08%, 0.12%, 0.16%, 0.20%, 0.60%, 0.80%, 1%, 5%, and 10% w/w. Partial least squares regression (PLSR) was utilized to predict the level of MP contamination. The results showed that MSI and fluorescence spectroscopy effectively predicted the four types of MP, resulting in R p 2 values ranging from 0.61 to 0.98 and root mean square error of prediction (RMSEP) values ranging between 0.43 and 2.89. On the other hand, partial least squares discriminant analysis (PLSDA) was employed to classify the contamination level of shrimp into three groups: low, medium, and high concentrations. PLSDA demonstrated exceptional classification accuracy ranging from 70.37% to 98.15% for all MP contamination using the three techniques (MSI, MIR, and fluorescence spectroscopy). These findings emphasize the effectiveness of spectroscopic techniques in detecting and quantifying microplastic contamination in shrimp, serving as a valuable tool for ensuring food safety and addressing environmental concerns associated with microplastics in the seafood supply chain. • MSI, MIR, and fluorescence predicted and classified 1−4 mm MPs in minced shrimp • PLSR predicted MP contamination from 0% to 10% w/w in shrimp samples • PLSDA classified MP contamination into low, medium, and high concentration groups • PLSDA classified four MP types (PET, PE, PP, and PS) with 70.37%−98.15% accuracy • PLSR predicted MPs using MSI and fluorescence, with R² of 0.61−0.98, RMSEP 0.43−2.89