Prediction of Tenderness, Juiciness, and Flavor Profile of 2 Beef Muscles with Different Aging Times Using Rapid Evaporative Ionization Mass Spectrometry (REIMS)

Abstract

Rapid evaporative ionization mass spectrometry (REIMS) is a novel technique that provides rapid chemical information on biological tissues and has the potential to predict beef quality attributes in real time. This study aims to assess the ability of analysis by REIMS coupled with chemometric modeling to predict the quality attributes of wet-aged beef at the grading time. USDA Select and upper two-thirds Choice (n=42, N=84) striploins (longissimus lumborum [LL]) and tenderloins (psoas major [PM]) were collected 36 h postmortem from a commercial beef abattoir. The LL and PM were cut into portions and aged for 3, 14, and 28 d. Aged samples were analyzed for slice shear force,Warner-Bratzler shear force (WBS), and trained sensory panels (tenderness, juiciness, and flavor attributes), and results were used to categorize both LL and PM into binary classes. Additionally, slivers of the longissimus dorsi muscle between the 12th and 13th rib were collected during grading (36 h postmortem) and analyzed using REIMS. The REIMS data were used to build predictive models for tenderness, juiciness, and flavor classes for the 3 aging periods and 2 muscles. Prediction accuracies of all models were higher than classifying the samples by chance (P<0.05), except WBS of 3 d aging model (P>0.05). However, model accuracies were not too high, which could be due to overlaps between classes, small sample sizes, and unbalanced data, which could negatively affect predictive models. Results demonstrated that the chemical finger-prints obtained with REIMS could potentially sort carcasses by flavor, juiciness, and tenderness in real time. However, the full realization of this approach will require an increased sample size and the development of a sampling method that allows improved separation between sensory classes.