Introduction
In South Korea, since 2007, pig carcass grading standards have been determined by categorizing them into primary meat quantity and secondary meat quality criteria, and then assigning a single grade that combines both meat quantity and quality (KLEI, 2009; Bae et al., 2016). Secondary evaluation is conducted by trained assessors who classify based on appearance criteria (such as fat content and pork belly condition), meat quality criteria (including fat deposition, meat color, meat texture, fat color, and fat content), and defect criteria (such as lesions, fractures, and muscle hemorrhage). However, currently, there isn't a significant difference in pork quality based on grades, and pig carcass grade represents meat quantity rather than meat quality. To address this issue, evaluating pork quality using traditional methods based on physical dissection adds costs, increases manpower demand, consumes time, and may involve occasional subjectivity based on the expertise of evaluators (Monziols et al., 2006; Prieto et al., 2009).
With the recent increase in consumption of grilled pork belly and pork shoulder, there is a growing demand for quality labeling of meat cuts. However, these preferred cuts, pork shoulder butt, and pork belly, are not exposed during grading process, making it difficult for experts to assess meat quality solely by visual inspection.
In recent years, with the advancement in technologies, several non-invasive and non-destructive techniques such as, spectroscopic techniques (e.g., Raman spectroscopy), imaging techniques (including radiography, and thermal imaging), multiplanar imaging techniques (such as magnetic resonance imaging and computed tomography), etc. have been introduced to study composition of carcasses or live animals and assess their quality (Wu et al., 2022). However, MRI (magnetic resonance imaging) has time constraints, and ultrasound is associated with the disadvantage of relatively low accuracy (Sammak et al., 1999). Therefore, CT (computed tomography) is considered to most suitable method to determine the level of muscle and fat content in a carcass.
While studies have been conducted to address these issues by examining the correlation between the entire carcass and the pork belly quality or to confirm the correlation between weight or genetics and the pork belly quality (Duziński et al., 2015; Lee et al., 2018; 2023), there is a noticeable lack of research on the correlation between pork belly and the pork shoulder butt. Therefore, in this review, we aimed to investigate the correlation between the muscle-to-fat ratio of pork belly and pork shoulder butt using CT.
Materials and Methods
Animals, slaughter, and dissection
The present study was conducted on 26 crossbred (Duroc × [Landrace × Large White]) gilt carcasses with an average live weight of 115 kg. The animals were slaughtered following the standard procedures of the Korea Institute for Animal Products Quality Evaluation at a slaughterhouse approved for the humane management and use of animals. The left half carcass was used in this study. At 24 h after slaughter, the pork shoulder butt and pork belly were dissected from the carcass for imaging examination.
Computed tomography
Each of the 26 pork belly and pork shoulder butt were scanned using a 32-detector-row CT scanner (Alexion™, Toshiba, Japan) with following scan parameters: 120 kilovoltage peak (kVp), 150 milliampere-seconds (mAs), 1 mm slice thickness, 0.75 second rotation time, and a 0.938 collimation beam pitch. The acquired CT images displaying a soft tissue window (window level = 40 HU (Hounsfield units), window width = 400 HU) were extracted using commercially available software (Xelis, INFINITT Healthcare Co., Ltd., Korea).
Image & statistical analysis
The image obtained from CT scans was checked through the picture archiving and communications system (PACS). The volume of muscle and fat in the pork belly and shoulder butt of cross-sectional images taken by CT was estimated using Vitrea workstation version 7 (Vital Images, USA). To calculate the muscle and fat content within the pork belly and shoulder butt, areas with the corresponding HU values on the CT scans were primarily identified manually. While the Vitrea software aided with semi-automated detection of contiguous regions exhibiting similar HU values, manual adjustments were frequently made to exclude any undesired areas. These selected regions were aggregated from each slice. Subsequent to this, the gathered data was processed using the Vitrea post-processing software to determine the volumes, which were measured in milliliters (mL). In addition to volume calculations, a three-dimensional reconstruction of the organ of interest was generated. Pearson's correlation coefficient was analyzed to evaluate the relationship by region (pork belly and shoulder butt), and statistical processing was conducted using GraphPad Prism 8 (GraphPad Software, USA).
Results and Discussion
Estimates of muscle and fat contents in the cuts
During the CT scan of the pork tissue (belly and shoulder butt), the absorption rate, which varies depending on muscle and fat density, was represented as a 3D image expressed in HU using reconstruction technology and displayed in Fig. 1 (Fonti-Furnols et al., 2009; Lucas et al., 2017; Xiong et al., 2017). After CT imaging and automated calculation using the Vitrea workstation, the muscle-to-fat ratio for pork belly was 1 : 0.86, and for pork shoulder, it was 1 : 0.37 (Table 1). As a result of CT analysis of the correlation coefficient between pork belly and pork shoulder butt in relation to the muscle-fat ratio, the correlation coefficient was 0.5679 (R2 = 0.3295, p < 0.01) (Fig. 2).
Non-invasive methods for assessing tissue composition without cutting the body include CT, MRI, and ultrasound. However, MRI has time constraints and ultrasound has the disadvantage of relatively low accuracy, so CT is considered to be the most appropriate method to determine the level of muscle and fat content in a carcass. CT, an important tool used in medical diagnosis, has gained significance in recent years for understanding the composition of carcasses and various livestock species such as sheep, goats, pigs, cattle, and more (Bunger et al., 2011). Researchers have reported using CT scanning to measure lean percentage and fat content in whole pig bodies and/or carcasses (Horn et al., 1997; Dobrowolski et al., 2004; Romvári et al., 2006). According to Dobrowolski et al. (2004), the CT method is a recommended reference method for determining meat content. Romvári et al. (2006) reported a correlation of 0.97 between muscle mass determined by CT scans and slaughter results measured using traditional methods, as well as a correlation of 0.95 for fat content.
The application of CT in this study allows for the non-invasive assessment of tissue composition in live animals. However, there are several key considerations that need to be addressed for widespread use. In post-mortem research, tracking over time is essential due to significant physicochemical alterations resulting from slaughter and post-mortem muscle water loss and fat crystallization. Additionally, Trusell et al. (2011) reported significant variations in muscle and fat distribution in pork belly based on anatomical vertebral positions. Hence, estimations of fat content through CT imaging still require improvement, especially with regard to the ratio between muscle and fat.
Conclusion
This study was conducted to investigate the correlation between meat quality and muscle fat content in pork cuts (pork belly and pork shoulder butt) not exposed in the carcass state using CT imaging techniques. The application of CT, as utilized in this study, allows for the non-invasive assessment of tissue composition in half carcass, and as a result, CT images provided very accurate estimates of muscle content. Therefore, the results of this study are expected to provide objective data that can be used as a reference for predicting and grading carcass quality. However, further, more detailed investigations are needed to validate these results.
Acknowledgements
This work was carried out with the support of “Cooperative Research Program for Agriculture Science and Technology Development (Project No. PJ0162112023, RS-2021-RD010001)” Rural Development Administration, Republic of Korea.
Authors Information
Sheena Kim, https://orcid.org/0000-0002-5410-1347
Jeongin Choi, https://orcid.org/0000-0001-9882-2368
Eun Sol Kim, https://orcid.org/0000-0001-8801-421X
Gi Beom Keum, https://orcid.org/0000-0001-6006-9577
Hyunok Doo, https://orcid.org/0000-0003-4329-4128
Jinok Kwak, https://orcid.org/0000-0003-1217-3569
Sumin Ryu, https://orcid.org/0000-0002-1569-3394
Yejin Choi, https://orcid.org/0000-0002-7434-299X
Sriniwas Pandey, https://orcid.org/0000-0002-6947-3469
Na Rae Lee, https://orcid.org/0009-0003-7230-9891
Jooyoun Kang, https://orcid.org/0000-0002-3974-2832
Yujung Lee, https://orcid.org/0000-0002-5939-4441
Dongjun Kim, https://orcid.org/0000-0001-8649-3157
Kuk-Hwan Seol, https://orcid.org/0000-0002-0907-882X
Sun Moon Kang, https://orcid.org/0000-0003-3947-4337
In-Seon Bae, https://orcid.org/0000-0003-3543-8785
Soo-Hyun Cho, https://orcid.org/0000-0002-8073-8771
Hyo Jung Kwon, https://orcid.org/0000-0001-5927-1970
Samooel Jung, https://orcid.org/0000-0002-8116-188X
Youngwon Lee, https://orcid.org/0000-0003-3207-0989
Hyeun Bum Kim, https://orcid.org/0000-0003-1366-6090
