Prediction of lamb carcass composition and meat quality using combinations of post-mortem measurements

Assessing the Feasibility of Using Kinect 3D Images to Predict Light Lamb Carcasses Composition from Leg Volume

This study aimed to evaluate the accuracy of the leg volume obtained by the Microsoft Kinect sensor to predict the composition of light lamb carcasses. The trial was performed on carcasses of twenty-two male lambs (17.6 ± 1.8 kg, body weight). The carcasses were split into eight cuts, divided into three groups according to their commercial value: high-value, medium value, and low-value group. Linear, area, and volume of leg measurements were obtained to predict carcass and cuts composition. The leg volume was acquired by two different methodologies: 3D image reconstruction using a Microsoft Kinect sensor and Archimedes principle. The correlation between these two leg measurements was significant (r = 0.815, p < 0.01). The models to predict cuts and carcass traits that include leg Kinect 3D sensor volume are very good in predicting the weight of the medium value and leg cuts (R² of 0.763 and 0.829, respectively). Furthermore, the model, which includes the Kinect leg volume, explained 85% of its variation for the carcass muscle. The results of this study confirm the good ability to estimate cuts and carcass traits of light lamb carcasses with leg volume obtained with the Kinect 3D sensor.

Effect of feeding Sesbania sesban and reed grass on growth performance, blood parameters, and meat quality of growing lambs

This study was conducted to evaluate the effects of green fodders, fresh Sesbania sesban, and reed grass, on growth performance, blood metabolites, and meat quality of growing lambs. Twenty-one male Farafra lambs (19.0 ± 1.87 kg body weight) were randomly distributed into three groups, each with seven animals. The first group was fed on 60% concentrates and 40% fresh alfalfa, and served as a control. The second and third groups were fed also on 60% concentrates plus 40% fresh Sesbania sesban and reed grass, respectively. Lambs were fed the diets until reaching the target slaughter weight of about 50.0 kg, which was obtained after a period of 215 days. Five lambs from each group were randomly selected and slaughtered at the end of the experiment. Slaughter body weight, carcass traits weight, and dressing percentage were estimated. Chemical composition and physical properties of longissimus dorsi muscle, from the animals' right side, were determined. It was found that no significant (P > 0.05) differences among groups in body weight, average daily gain, feed conversion rate, hot carcass, dressing percentage, and carcass components were observed. The forage intake of Sesbania and alfalfa tended to be higher than that of reed forage. There were no significant (P > 0.05) differences in the blood parameters among the treatments, except the urea level which was higher in Sesbania sesban group than the reed one. However, the serum urea concentration in Sesbania and reed groups was similar to that of alfalfa. Sesbania sesban and reed forages decreased (P < 0.05) the eye muscle area and shear force in lambs when compared to the control group. The longissimus dorsi muscle of the Sesbania and reed treatments was found to be numerically higher in the moisture and lower in the fat and collagen contents than the alfalfa group. Most of meat and fat color parameters were improved by feeding of Sesbania and reed forages. In conclusion, Sesbania sesban or reed grass, as unconventional forage sources, can be used as alternatives of the commonly used alfalfa feed. Moreover, these sources could enhance the meat quality of lambs and increasing the tenderness with accepted meat and fat color. So, it can be recommended to use Sesbania or reed grass forages in diets of sheep, to overcome the rising in the price or shortage in the availability of traditional green fodders.

Evaluation of the functionality of EUROP carcass classification system in thin-tailed and fat-tailed lambs

Aim of study was to investigate the relationship of EUROP carcass fatness and conformation classes with carcass and meat quality characteristics in thin-tailed (Kivircik) and a fat-tailed (Kangal Akkaraman) lambs. A total of 96 lamb carcasses (48 Kivircik and 48 Kangal Akkaraman breed) were examined. Distribution of carcasses in conformation and fatness classes were more limited in Kangal Akkaraman carcasses. More accurate predictions were obtained in Kivircik lambs compared to Kangal Akkaraman lambs in terms of backfat thickness, kidney knob and channel fat percentage, longissimus thoracis muscle area, weights of high-value carcass joints, weights of tissues in the hind limb and muscle/bone ratio. EUROP classification system was not appropriate for predicting meat quality in thin-tailed and fat-tailed lambs. The combination of EUROP fatness and conformation classes with cold carcass weight can be used to predict weights of high-value carcass joints and weights of hind limb tissues.

Predicting the chemical composition of the body and the carcass of hair sheep using body parts and carcass measurements

Determination of the chemical composition in the body and carcass of ruminants is important for both nutritional requirement studies and the meat industry. This study aimed to develop equations to predict the body and carcass chemical composition of hair sheep using the chemical composition of body parts, carcass measurements and shrunk BW as predictors. A database containing 107 individual records for castrated male hair sheep ranging from 24 to 43 kg BW was gathered from two body composition studies. The empty body, carcass and body parts were analyzed for water, ash, fat and protein contents (%). The body parts used to estimate body and carcass composition were fore leg, hind leg and 9-11th rib section. The carcass measurements used were leg length, thoracic circumference, hind circumference, hind width, thoracic width, thoracic depth and chest width. Each model performance was evaluated using a leave-one-out cross-validation. Multiple regression analysis considering the study as a random effect revealed that body parts in association with carcass measurements were significant for predicting the chemical composition in the body of castrate male sheep. However, the use of the chemical composition of hind leg produced the best models for predicting the ash and fat contents in the empty body, whereas the water and protein contents in the empty body were better predicted when using the chemical compositions of 9-11th rib section and fore leg, respectively. Multiple regression analysis also revealed that most body parts were suitable for predicting the carcass composition, except for 9-11th rib section whose chemical composition did not produce significant prediction equations for ash and protein carcass contents. The use of the chemical composition of hind leg in association with carcass measurements produced the best models for predicting the water and fat contents in the carcass, while the ash and protein contents in the carcass were better predicted when using the chemical composition of fore leg. In conclusion, precision, accuracy and goodness-of-fit of the equations drove the selection of the chemical composition of hind leg and carcass measurements in a multivariate approach, as the most suitable predictors of the chemical composition of the body and carcass of hair sheep. However, the chemical composition of fore leg may be used as well. The developed equations could improve the accuracy of the empty body and carcass composition estimations in sheep, optimizing the estimation of nutrient requirements, as well as the carcass quality evaluation for this species.

Comparison of the decision tree, artificial neural network and multiple regression methods for prediction of carcass tissues composition of goat kids

The aim of this study was to predict carcass tissue composition of goat kids using the decision tree with CHAID algorithm (DT) and artificial neural network (ANN) method in comparison with classical step-wise regression (SWR) analyse. Data were obtained from 57 goat kids of Gokceada breed. Predictor variables were pre-slaughter weight, several carcass measurements and indices, weights of different carcass joints and dressing percentage. R² values ranging from 0.212 to 0.371 indicating low to moderate accuracy were obtained for predicting muscle proportion. DT and ANN yielded similar R² values for predicting bone proportion. DT was the best prediction method for estimating proportions of subcutaneous fat (R² = 0.828) and intermuscular fat (R² = 0.789). According to DT analyses, cold carcass weight was the most important factor influencing bone proportion, while kidney knob and channel fat weight was the predominant factor influencing subcutaneous, intermuscular and total fat proportions. Consequently, the use of DT method can be considered to predict carcass fat proportions.

Preliminary estimation of fat depth in the lamb short loin using a hyperspectral camera

The objectives of the present study were to describe the approach used for classifying surface tissue, and for estimating fat depth in lamb short loins and validating the approach. Fat versus non-fat pixels were classified and then used to estimate the fat depth for each pixel in the hyperspectral image. Estimated reflectance, instead of image intensity or radiance, was used as the input feature for classification. The relationship between reflectance and the fat/non-fat classification label was learnt using support vector machines. Gaussian processes were used to learn regression for fat depth as a function of reflectance. Data to train and test the machine learning algorithms was collected by scanning 16 short loins. The near-infrared hyperspectral camera captured lines of data of the side of the short loin (i.e. with the subcutaneous fat facing the camera). Advanced single-lens reflex camera took photos of the same cuts from above, such that a ground truth of fat depth could be semi-automatically extracted and associated with the hyperspectral data. A subset of the data was used to train the machine learning model, and to test it. The results of classifying pixels as either fat or non-fat achieved a 96% accuracy. Fat depths of up to 12 mm were estimated, with an R2 of 0.59, a mean absolute bias of 1.72 mm and root mean square error of 2.34 mm. The techniques developed and validated in the present study will be used to estimate fat coverage to predict total fat, and, subsequently, lean meat yield in the carcass.

The value of objective online measurement technology: Australian red meat processor perspective

In the past, the adoption of online measurement technologies for measuring carcass and meat quality traits objectively has been low among Australian red meat processors. The aim of the present work was to obtain a greater understanding of Australian processor views on the value of objective online measurement technologies. This was achieved through consultation with 65 Australian processors, to understand which carcass and meat quality traits they considered important to objectively measure and what they thought of current and future technologies. It was shown that beef processors ranked meat colour and tenderness as the most important traits (P < 0.001) to objectively measure online. Sheep processors ranked tenderness, pH, age, meat colour, total tissue depth at the 12th rib 110 mm from the midline (GR) and saleable meat yield percentage as the most important traits (P < 0.001) to objectively measure online. The overall processor responses indicated that there is support for online measurement technologies, with 80% of processors stating that online objective grading systems have a role in the Australian meat processing sector now and 88% considered these to have a role in the future. Much can be learned from the implementation of previous online objective measurement technologies by processors in terms of commercialisation and adoption strategies. The development and adoption of objective online measurement technologies is challenging and complex. However, increased adoption of online measurement technologies has the potential to achieve benefits to the whole of industry and needs continued support, coupled with new approaches to enhance adoption.

Prediction of foal carcass composition and wholesale cut yields by using video image analysis

This work represents the first contribution for the application of the video image analysis (VIA) technology in predicting lean meat and fat composition in the equine species. Images of left sides of the carcass (n=42) were captured from the dorsal, lateral and medial views using a high-resolution digital camera. A total of 41 measurements (angles, lengths, widths and areas) were obtained by VIA. The variation of percentage of lean meat obtained from the forequarter (FQ) and hindquarter (HQ) carcass ranged between 5.86% and 7.83%. However, the percentage of fat (FAT) obtained from the FQ and HQ carcass presented a higher variation (CV between 41.34% and 44.58%). By combining different measurements and using prediction models with cold carcass weight (CCW) and VIA measurement the coefficient of determination (k-fold-R 2) were 0.458 and 0.532 for FQ and HQ, respectively. On the other hand, employing the most comprehensive model (CCW plus all VIA measurements), the k-fold-R 2 increased from 0.494 to 0.887 and 0.513 to 0.878 with respect to the simplest model (only with CCW), while precision increased with the reduction in the root mean square error (2.958 to 0.947 and 1.841 to 0.787) for the hindquarter fat and lean percentage, respectively. With CCW plus VIA measurements is possible to explain the wholesale value cuts yield variation (k-fold-R 2 between 0.533 and 0.889). Overall, the VIA technology performed in the present study could be considered as an accurate method to assess the horse carcass composition which could have a role in breeding programmes and research studies to assist in the development of a value-based marketing system for horse carcass.

Post-mortem prediction of primal and selected retail cut weights of New Zealand lamb from carcass and animal characteristics

Post-mortem measurements (cold weight, grade and external carcass linear dimensions) as well as live animal data (age, breed, sex) were used to predict ovine primal and retail cut weights for 792 lamb carcases. Significant levels of variance could be explained using these predictors. The predictive power of those measurements on primal and retail cut weights was studied by using the results from principal component analysis and the absolute value of the t-statistics of the linear regression model. High prediction accuracy for primal cut weight was achieved (adjusted R(2) up to 0.95), as well as moderate accuracy for key retail cut weight: tenderloins (adj-R(2)=0.60), loin (adj-R(2)=0.62), French rack (adj-R(2)=0.76) and rump (adj-R(2)=0.75). The carcass cold weight had the best predictive power, with the accuracy increasing by around 10% after including the next three most significant variables.

Evaluation of Columbia, U.S. Meat Animal Research Center Composite, Suffolk, and Texel rams as terminal sires in an extensive rangeland production system: VI. Measurements of live-lamb and carcass shape and their relationship to carcass yield and value

Linear measurements on live lambs and carcasses can be used to characterize sheep breeds and may have value for prediction of carcass yield and value. This study used 512 crossbred lambs produced over 3 yr by mating Columbia, U.S. Meat Animal Research Center (USMARC) Composite, Suffolk, and Texel rams to adult Rambouillet ewes to assess sire-breed differences in live-animal and carcass shape and to evaluate the value of shape measurements as predictors of chilled carcass weight (CCW), weight of high-value cuts (rack, loin, leg, and sirloin; HVW), weight of trimmed high-value cuts (trimmed rack and loin and trimmed, boneless leg and sirloin; TrHVW), and estimated carcass value before (CVal) and after trimming of high-value cuts (TrCVal). Lambs were produced under extensive rangeland conditions, weaned at an average age of 132 d, fed a concentrate diet in a drylot, and harvested in each year in 3 groups at target mean BW of 54, 61, and 68 kg. Canonical discriminant analysis indicated that over 93% of variation among sire breeds was accounted for by the contrast between tall, long, less-thickly muscled breeds with greater BW and CCW (i.e., the Columbia and Suffolk) compared with shorter, more thickly muscled breeds with smaller BW and CCW. After correcting for effects of year, harvest group, sire breed, and shipping BW, linear measurements on live lambs contributed little to prediction of CCW. Similarly, after accounting for effects of CCW, linear measurements on live animals further reduced residual SD (RSD) of dependent variables by 0.2 to 5.7%, with generally positive effects of increasing live leg width and generally negative effects of increasing heart girth. Carcass measurements were somewhat more valuable as predictors of carcass merit. After fitting effects of CCW, additional consideration of carcass shape reduced RSD by 2.1, 3.6, 9.5, and 2.2% for HVW, TrHVW, CVal, and TrCVal, respectively. Effects of increasing carcass leg width were positive for HVW, TrHVW, and TrCVal. We also observed positive effects of increasing carcass length on TrCVal and negative effects of increasing cannon bone length on HVW and CVal. Increasing shoulder width had positive effects on CVal but negative effects on TrHVW. Differences in lamb and carcass shape were significantly associated with carcass yield and value, but the additional accuracy associated with use of these measurements was modest relative to that achieved from use of only shipping BW or CCW.

Effects of a quantitative trait locus for increased muscularity on carcass traits measured by subjective conformation and fat class scores and video image analysis in crossbred lambs

A quantitative trait locus (QTL) for increased loin muscularity (TM-QTL) has previously been identified in purebred Texel sheep. Crossbred lambs born out of Mule ewes mated to heterozygous Texel sires for the TM-QTL were evaluated for a range of carcass traits. Lambs were genotyped and classified as carriers (n = 62) of a single copy of the TM-QTL and non-carriers (n = 49). In this study, the effects of the TM-QTL on carcass attributes were investigated using subjective classification scores for conformation and fatness, and measurements from a video image analysis (VIA) system. In addition, refined prediction equations to estimate weights of primal joints (leg, chump, loin, breast and shoulder) were obtained by calibrating the VIA system against computer tomography (CT) measurements in the loin region. The new refined prediction models increased the accuracy of prediction of all primal cuts on an average of 16% compared to previously derived standard VIA prediction equations. The coefficient of determination (R2) of the VIA system to predict in vivo CT measurements ranged from 0.39 to 0.72 for measurements of Musculus longissimus lumborum (MLL) area, width and depth, lumbar spine length, loin muscle volume and loin muscularity index. Using VIA estimates of CT-measured loin muscle traits, a significant increase in depth (+2.7%) of the MLL was found to be associated with the TM-QTL. Conformation and fatness scores and the shape of the carcass measured as individual lengths, widths and areas by VIA were not significantly influenced by the TM-QTL. Primal meat yields estimated using both standard and refined VIA prediction equations were not significantly affected by the TM-QTL. However, carcass 'compactness' was found to have significantly increased in carrier lambs. The weight of the dissected MLL estimated using VIA information was greater (+2.6%) for carriers compared to non-carriers. To conclude, neither the current industry carcass evaluation system for conformation and fatness nor the standard VIA system is able to identify the effect of the TM-QTL in the loin region in the moment. However, the calibration of the VIA system against CT measurements resulted in improved VIA prediction equations for primal meat yields and also showed moderate potential to estimate loin muscle traits measured by CT and to detect, partially, the effect of the TM-QTL on these traits.