Monitoring the freshness change of stored Pacific whiteleg shrimp through inkjet-printed composite indicator coupled with artificial neural network

Pacific whiteleg shrimp ( Litopenaeus vannamei ) is a significant aquatic product extensively produced in the city of Zhuhai, Guangdong and other regions in the Greater Bay Area of China. Due to its high perishability during post-harvest storage, deteriorative reactions can result in substantial quality decrease over time, thereby causing potential economic loss and food wastage in the fishery industry. Hence developing a supplementary method in real-time monitoring shrimp freshness change is promising. In this study, we developed an inkjet-printed composite indicator, which was based on printing three pre-formulated inks onto photo papers: bromocresol purple, indigo carmine red, and reduced indigo carmine. The color-responsive indicator was applied inside the package to monitor the freshness change of whiteleg shrimps stored at 4 ºC for 9 days, where chemical and microbiological analyses were also conducted to assess shrimp quality. It was observed that the color changes of the indicator over storage were well correlated with the shrimp quality parameters including pH, total volatile basic nitrogen (TVB-N), total microbial counts and lipid oxidation products. After processing the indicator’s color response images, the artificial neural network (ANN) models were constructed, demonstrating strong predictive capability, especially with R 2 over 0.85 for pH and peroxide value. These positive results showed that the developed colorimetric indicator coupled with ANN can be a potential non-destructive method in evaluating real-time shrimp freshness, which may be also applicable to other fish and shellfish products with similar spoilage mechanisms. • Composite colorimetric indicator was developed via piezoelectric inkjet printing. • Indicator’s color response was substantial during shrimp deterioration over 9 days. • Biochemical changes and microbial growth of stored shrimp were analyzed. • Artificial neural network was employed to well predict real-time shrimp quality.