Greek ruminant meat production is challenged by the lack of objective carcass and meat evaluation protocols. X-ray Computed tomography (CT) is used as an innovative tool for carcass evaluation in many farm animal species. For the first time in Greece, in 2 beef, 13 sheep and 6 goat carcasses CT scans were performed to estimate carcass traits and composition. Image analysis protocols, formerly implemented in other species, were used to estimate carcass length and width as well as fat, muscle and bone volumes and weights. Our preliminary results indicate that accurate carcass composition estimations might be possible with the use of CT and image analysis. Results should be reevaluated on a bigger sample size. Concurrent carcass dissections could facilitate the validation of CT estimations. Standardization of carcass composition evaluation protocols based on non-destructive methods can certify meat quality and contribute to the overall competitiveness and sustainability of the sector.
Ruminant meat production in Greece represents 57.28% of non-poultry meat
production. In 2018, Greek beef, sheep and goat annual meat productions were 39.65,
50.57 and 19.56 thousand tons, respectively
So far, CT implementation on sheep aimed to predict carcass composition on live
animals mostly for breeding purposes. Cross-sectional scans at three to seven specific
anatomical positions were performed; non-carcass tissues were segmented
semiautomatically with a specially designed image analysis software
An approach based on spiral CT scanning of primal cuts and image analysis with
special software has been used to estimate beef carcass composition
In another approach implemented on pigs, spiral CT scans are obtained from half
carcasses or specific commercial cuts and image analysis software is used to separate
tissues. To estimate tissue volumes and weights, voxel dimensions and tissue densities
are utilized; results are highly accurate
Considering the strengths and weaknesses of the above methods and the peculiarities of Greek ruminant meat production, the present study is a preliminary approach of CT as a post-mortem carcass evaluation tool in Greece.
The objective was twofold; (i) to use CT and image analysis protocols designed for carcass evaluation in other species (ie. pigs), to estimate sheep, goat and beef carcass traits (length, width) and composition parameters (volume and weight of muscle, fat and bone tissues) and (ii) to compare the estimated sheep carcass quality parameters with respective dissection data of a previous study. 3
Dairy sheep and goats (mostly fat-tailed Chios sheep or other crossbreds) of both sexes at different live weights (representing 25%, 35%, 50%, 70% and 100% of mature weight –Table 1) were selected. Male, crossbred beef cattle at the optimum finishing weight were also selected. Animals were slaughtered in three commercial slaughterhouses. Sheep and goat carcasses were chilled for 24 hours then transferred to the CT scanner located at the Laboratory of Diagnostic Imaging, School of Veterinary Medicine, Aristotle University of Thessaloniki. Chilled beef carcasses were scanned 48 hours after slaughtering. In total, 13 sheep, 6 goat and 2 beef carcasses were scanned.
Beef carcasses were split in halves and one half was scanned. Beef carcass halves were dissected into six primal cuts (Figure 1) to facilitate the scanning procedure. Sheep and goat carcasses were scanned intact and contained remaining organs (heart, lungs, liver, internal fat, kidneys), except for heavier carcasses (70% and 100% of mature weight). The latter were longer than the maximum scanning length (110 cm); thus, two scans were performed representing front and back halves of carcasses. A helical volume of data comprising the carcass was acquired at 150 mAs and 120 kV, acquisition matrix 512x512 and convolution kernel STANDARD (soft tissues) using a 16-row multi-detector CT scanner (Optima CT520, GE Hangwei Medical Systems, Beijing China) (Figure 2). Transverse overlapping slices of carcass were obtained. Slice thickness for small ruminants was 0.625 mm and ranged from 0.625 to 3.75 mm for beef. Field of view (FoV) ranged depending on carcass width, as did tube current since dose efficiency parameter (Optidose) was active (Table 2).
Remaining organs segmentation was performed using TurtleSeg software (version 1.2.1). To eliminate differences due to operator effect, only one trained researcher performed the protocol. Viscera area was selected by semi-manual contouring in several images (Figure 3); after completion, a 3-dimensional grid containing all viscera was created and exported as a new set of images, which was used to subtract these organs from the original dataset.
Sheep carcass dissection data (muscle and fat weight) of an earlier study
4.1
Means of carcass traits and composition parameters for each live weight category per species are presented in Tables 3a and 3b. Total number of carcasses differs in each weight category and in some cases no carcasses have been examined so far. As the study is still ongoing, results presented here are preliminary. Completion of the experiment will provide a larger sample of carcasses equally distributed in each live weight category within species.
Estimated carcass length and width were in the normal range for each species. The method was fast and easily applicable; it is a good alternative to manual measurements.
As discussed in the introduction, CT scanning of sheep has so far been implemented in variable experimental protocols and on live animals. The latter complicate a direct comparison of results. Anderson et al. (2015), used a similar approach on a large sample (n=1665) of Merino sheep (mean carcass weight: 23.3 kg); mean fat and muscle percentage were 27.0% and 57.1% of carcass weight, respectively. In the present study, the respective values were 16.7% and 40.5%. The difference may be due to variations between populations and samples; herein, a very limited sample (n=13) of dairy sheep carcasses was used and mean carcass weighted only 16.9 kg.
Studies on goats are limited and close to sheep ones regarding methods (Sørensen,
1992; Németh et al., 2010). In a recent study
Two beef carcasses (mean carcass weight: 389.9 kg) were examined and mean fat, muscle and bone percentages estimated were 12.0, 70.1 and 17.9%, respectively. Navajas et al. (2010a) followed a similar protocol for a bigger sample (n=44, mean carcass weight: 356.5 kg) and estimated mean fat, muscle and bone percentage of 20.4, 64.1 and 15.4%. Except for fat percentage, the other estimations are quite similar considering the differences between the two datasets.
The present results compared to other studies, demonstrate a tendency towards lower estimated fat and fluctuating muscle weights. Variable experimental designs and population parameters among studies complicate safe assumptions. Lack of dissection data directly related to the present dataset does not allow the proper validation of results. Carcass Traits beef
Means of fat and muscle weight of dissected sheep carcasses are presented in Table 4. Mean carcass weights for each live weight category are similar between the two datasets; percentage change of mean CT-estimated values compared to dissection data varied from 1.71 to 4.53% by absolute value. Regarding fat weight means, slight differences were observed (percentage change: 3.24 – 6.23% by absolute value), except for light carcasses (25% of adult weight). Muscle weight means were slightly different for middleweight categories (35 and 50% of adult weight – percentage change: 3.96 – 7.74% by absolute value). In contrast, percentage changes were larger for heavier and lighter sheep (10.92 – 17.68% by absolute value). Generally, CTestimated muscle and fat weights were close to dissection data regarding middleweight carcasses. Fat weight estimation was more uniform compared to muscle (except for light carcasses). A variety of factors, such as genetic differences and dissection quality, may be causing the observed discrepancies. beef beef sheep 25% sheep 35% sheep 50% sheep 100% sheep total fat weight (g) 1159.69 1947.80 2991.47 8608.19 3676.79 fat weight (g)
This is the first study to implement CT as a carcass evaluation tool in Greece. Preliminary results presented in this manuscript indicate potential in this field. Our next tasks focus on the enrichment of the sample with more CT-scanned carcasses and the inclusion of dissection data that will directly validate the results.
Standardization of a carcass evaluation protocol, based on non-destructive methods, can significantly change meat industry in Greece. Accurate and easy carcass quality evaluation will facilitate high-quality meat production, competitive against imported special cuts and able to reach high selling prices. Objective quality perception will permit grading of different meat cuts, thus allowing for better produce capitalization and waste minimization. This effort will form a value chain, rewarding quality meat producers, creating bigger profit margins and meeting consumer demands. Acknowledgments. This work was funded by GreQuM project (Code: T1EDK05479), co-financed by Greece and EU through EPAnEK 2014-2020 and Partnership Agreement 2014-2020.