An experimental protocol to determine quality parameters of dry-cured loins using low-field Magnetic Resonance Imaging

Abstract

The objective of this study was to achieve an experimental protocol (EP) to determine quality characteristics of dry-cured loins non-destructively by using low-field (LF) Magnetic Resonance Imaging (MRI). The MRI procedure is composed of three main stages: MRI acquisition, MRI analysis (computer vision techniques) and data analysis (data mining methods). Two procedures have been implemented within a EP and validated with real samples from the meat industry (dry-cured loins, n = 100) by means of different quality measures. The validation results may indicate the use of both implemented procedures and the development of an EP to determine quality characteristics of loins by LF MRI-computer vision-data mining in a non-destructive way, with high accuracy and reducing the dispersion of the values. This brings the possibility of implementing this methodology in meat processing plants. • MRI with different computer vision and data mining methods allow the analysis of meat products. • Two MRI - computer vision - data mining procedures was implemented and evaluated. • An experimental protocol was developed to analyse dry-cured loins non-destructively.