Determination of decomposition in Salmon products by mass spectrometry with sensory‐driven multivariate analysis

Sensory analysis is performed in USFDA testing laboratories to determine the decomposition status of seafood samples. Current alternatives are not always directly comparable to sensory testing. In this study, a novel method was explored using LC-HRMS with untargeted analysis and a statistical model driven by sensory results via a script written in the R programming language. This allows generation of data, which are directly comparable to commonly used sensory scoring techniques. The method was applied to six species of salmon, across a wide range of decomposition conditions. Accuracy and reproducibility of the method in comparison with sensory results were demonstrated and show promising results. Correlation values between calculated and sensory values for each species ranged from 0.912 to 0.973, and the correct status was determined for 96% of samples (n = 92). Since it generates data which are comparable to sensory analysis, it provides a potentially useful companion to that work. Practical Application Spoilage of seafood samples is currently tested in USFDA laboratories by sensory analysis to help protect consumer safety. This study aims to identify a chemistry-based analytical complementary method to that work for use in the regulatory environment.