A multi-block data approach to assessing beef quality: ComDim analysis of hyperspectral imaging,
Qian You1, Ziyuan Wang1, Xingguo Tian1
1Guangdong Provincial Key Laboratory of Food Quality and Safety, Nation-Local Joint Engineering Research Center for Machining and Safety of Livestock and Poultry Products, South China Agricultural University, Guangzhou 510642, China.
Food Chemistry
|June 4, 2023
Summary
This study applied ComDim, a multi-block data analysis method, to assess beef quality using diverse data sources. ComDim effectively differentiated beef types by revealing relationships between techniques and quality metrics.
Area of Science:
- Food Science
- Chemometrics
- Analytical Chemistry
Background:
- Beef quality is influenced by numerous factors, necessitating advanced analytical techniques.
- Multi-block data analysis methods in chemometrics are valuable for integrating information from diverse sources.
- Hyperspectral imaging, NMR, electronic nose, and texture analysis provide complementary data for beef quality assessment.
Purpose of the Study:
- To apply the ComDim (Common Discriminant) method for multi-block data analysis of beef quality.
- To evaluate the efficiency of ComDim compared to traditional Principal Component Analysis (PCA) for data fusion.
- To differentiate beef quality attributes between tenderloin and hindquarter cuts using integrated analytical data.
Main Methods:
- Utilized ComDim, a multi-block data analysis technique, to integrate data from hyperspectral imaging, image texture, 1H NMR spectroscopy, quality parameters, and electronic nose.
- Employed Principal Component Analysis (PCA) as a benchmark for low-level data fusion.
- Analyzed beef samples from tenderloin and hindquarter cuts.
Main Results:
- ComDim demonstrated superior efficiency and power in analyzing multi-source data compared to PCA-based fusion methods.
- The study successfully differentiated beef tenderloin from hindquarter based on quality and metabolite composition.
- Key quality indicators like L* value and shear force were correlated with specific beef cuts, with tenderloin showing low L* and high shear.
Conclusions:
- ComDim is a powerful and efficient approach for characterizing samples by integrating data from multiple analytical techniques.
- The proposed strategy validates the utility of ComDim in assessing complex biological samples like beef.
- Multi-block analysis using ComDim provides a comprehensive understanding of beef quality variations across different cuts.
