Rapid Detection of Meat Varieties Based on Two-Dimensional Relaxation Fingerprint Spectroscopy

Objectives: Meat is one of the indispensable foods in human diet. Since the 21st century, the consumption of meat has been continuously increasing. However, current meat quality detection technologies commonly face several challenges, including complex sample pretreatment, long detection cycles, and high equipment costs, which significantly limit their practical effectiveness. To address these issues, this study has developed a novel detection system based on two-dimensional nuclear magnetic resonance (NMR) relaxation fingerprinting. Methods: By optimizing the low-field NMR experimental parameters—setting the waiting time between two sampling sequences D_0 = 500 ms and the inversion recovery time D_3 = 950 ms—two-dimensional relaxation fingerprint data were collected from 25 samples across five common meat varieties (duck, lamb, chicken, beef, and pork) and their various cuts. The study employed a three-dimensional characteristic parameter coordinate system for the quantitative characterization of relaxation fingerprints and constructed 80% confidence ellipsoid models for species identification. Additionally, sensitivity and linear fitting analyses were conducted on binary mixed samples with varying mutton-to-duck ratios (10%-90%). Results: The system offers notable advantages, including simple operation (no complex pretreatment required), rapid analysis (each test completed in under 15 minutes), and high identification accuracy. It effectively meets the detection requirements for major commercially available meat products. Conclusions: As a detection approach that combines theoretical innovation with practical feasibility, this technology not only provides a novel tool for food safety supervision but also shows great potential for application in quality control within the meat industry and in the standardization efforts of relevant testing agencies. It holds broad prospects for industrial implementation.