Online evaluation of a commercial video image analysis system (Computer Vision System) to predict beef carcass red meat yield and for augmenting the assignment of USDA yield grades. United States Department of Agriculture

Objective quantification of differences in wholesale cut yields of beef carcasses at plant chain speeds is important for the application of value-based marketing. This study was conducted to evaluate the ability of a commercial video image analysis system, the Computer Vision System (CVS) to 1) predict commercially fabricated beef subprimal yield and 2) augment USDA yield grading, in order to improve accuracy of grade assessment. The CVS was evaluated as a fully installed production system, operating on a full-time basis at chain speeds. Steer and heifer carcasses (n = 296) were evaluated using CVS, as well as by USDA expert and online graders, before the fabrication of carcasses into industry-standard subprimal cuts. Expert yield grade (YG), online YG, CVS estimated carcass yield, and CVS measured ribeye area in conjunction with expert grader estimates of the remaining YG factors (adjusted fat thickness, percentage of kidney-pelvic-heart fat, hot carcass weight) accounted for 67, 39, 64, and 65% of the observed variation in fabricated yields of closely trimmed subprimals. The dual component CVS predicted wholesale cut yields more accurately than current online yield grading, and, in an augmentation system, CVS ribeye measurement replaced estimated ribeye area in determination of USDA yield grade, and the accuracy of cutability prediction was improved, under packing plant conditions and speeds, to a level close to that of expert graders applying grades at a comfortable rate of speed offline.     

Effect of frame size, muscle score, and external fatness on live and carcass value of beef cattle.

Commercial slaughter steers (n = 329) and heifers (n = 335) were selected to vary in slaughter frame size and muscle thickness score, as well as adjusted 12th rib fat thickness. After USDA carcass grade data collection, one side of each carcass was fabricated into boneless primals/subprimals and minor tissue components. Cuts were trimmed to 2.54, 1.27, and .64 cm of external fat, except for the bottom sirloin butt, tritip, and tenderloin, which were trimmed of all fat. Four-variable regression equations were used to predict the percentage (chilled carcass weight basis) yield of boneless subprimals at different fat trim levels (.64, 1.27, and 2.54 cm) as influenced by sex class, frame size, muscle score, and adjusted 12th rib fat thickness. Carcass component values, total carcass value, carcass value per 45.36 kg of carcass weight, and live value per 45.36 kg of live weight were calculated for each phenotypic group and external fat trim level. Carcass fatness and muscle score had the most influence on live and carcass value (per 45.36 kg weight basis). Carcasses with .75 and 1.50 cm of fat at the 12th rib were more valuable as the trim level changed from 2.54 cm to .64 cm; however, for carcasses with 2.25 cm of fat at the 12th rib, value was highest at the 2.54 cm trim level. Value was maximized when leaner cattle were closely trimmed. There was no economic incentive for trimming light-muscled or excessively fat carcasses to .64 cm of external fat.     

Determining beef carcass retail product and fat yields within 1 hour postmortem.

Hot carcasses from 220 steers (progeny of Hereford or Angus dams mated to Angus, Charolais, Galloway, Gelbvieh, Hereford, Longhorn, Nellore, Piedmontese, Pinzgauer, Salers, or Shorthorn sires) were used to develop equations to estimate weights and percentages of retail product (RP) and trimmable fat (TF) yields. Independent variables examined were 1) 12-13th rib fat probe (12RFD), 2) 10-11th rib fat probe (10RFD), 3) external fat score (EFS), 4) percentage of internal fat estimated hot (H%KPH), 5) hindquarter muscling score (HQMS), and 6) hot carcass weight (HCW). Right sides of the carcasses were fabricated into boneless retail cuts, trimmed to .76 cm of subcutaneous and visible intermuscular fat, and weighed. Cuts were trimmed to 0 cm of subcutaneous and visible intermuscular fat and reweighed. Multiple linear regression equations containing 12RFD, EFS, H%KPH, and HCW accounted for 95 and 89% of the variation in weight of total RP at .76 and 0 cm of fat trim, respectively. When weights of RP from the four primal cuts (.76 and 0 cm of fat trim) were the dependent variables, equations consisting of 12RFD, EFS, H%KPH, and HCW accounted for 93 to 84% of the variation. Hot carcass equations accounted for 83% of the variation in weight of total TF at both .76 and 0 cm of fat trim. Furthermore, equations from hot carcass data accounted for 54 and 51% of the variation in percentage of total RP and 57 and 50% of the variation in percentage of RP from the four primal cuts at .76 and 0 cm of fat trim, respectively. Hot carcass prediction equations accounted for 72% of the variation in percentage of total TF at both fat trim levels. Hot carcass equations were equivalent or superior to equations formulated from chilled carcass traits.     

Yield grade and carcass weight effects on the cutability of lamb carcasses fabricated into innovative style subprimals.

Lamb carcass (n = 100) were selected from USDA yield grades (YG) 2, 3, and 4 and carcass weight (CW) groups 20.4 to 24.9, 25.0 to 29.5, and 29.6 to 34.0 kg. Lamb carcass were fabricated into semiboneless and boneless subprimals and trimmed to three s.c. fat trim levels: .64, .25, and .00 cm of fat remaining. Innovative subprimals were fabricated and yields were calculated for the subprimals and dissectible components (lean, bone, connective tissue, external fat, and seam fat) from each of the various subprimals. Carcass weight as a main effect in a two-way analysis of variance did not account for a significant amount of the variation in yield among trimmed subprimals or the percentage of the dissectible components, but USDA YG was a significant main effect in determining variation in yield for many of the subprimals or dissectible components. Muscle seaming of shoulders and legs and removal of excessive tails on the loin and rack resulted in a majority of the seam fat being removed from these cuts. Dissection data clearly showed that seam fat is a major component of rack and shoulder cuts and with increasing fatness or higher numerical yield grade there are clearly increased amounts of this depot. Increased trimming of external fat magnifies and draws more attention to the amount of seam fat remaining. Production of heavy, lean lambs would be more useful in an innovative type of program because of the larger-sized muscles. Heavy, fat lambs would not be as useful because of their decreased yields and excess seam fat located in cuts that cannot be muscled-seamed because of the loss of retail cut integrity. Seam fat was highly correlated to percentage of kidney and pelvic fat and to external fat thickness and with USDA yield grade but was not strongly correlated to carcass weight.     

Estimates of subprimal yields from beef carcasses as affected by USDA grades, subcutaneous fat trim level, and carcass sex class and type.

One hundred beef carcasses were selected to represent the mix of cattle slaughtered across the United States. Selection criteria included breed type (60% British/continental European, 20% Bos indicus, and 20% dairy carcasses), sex class (beef and Bos indicus: 67% steers, 33% heifers; dairy: 100% steers), USDA quality grade (4% Prime, 53% Choice, and 43% Select), USDA yield grade (10% YG 1, 43% YG 2, 40% YG 3, and 7% YG 4), and carcass weight (steers: 272.2 to 385.6 kg, heifers: 226.8 to 340.2 kg). One side of each carcass was fabricated into boneless subprimals and minor cuts following Institutional Meat Purchase Specifications. After fabrication, subprimals were trimmed progressively of fat in .64-cm increments beginning with a maximum of 2.54 cm and ending with .64 cm. Linear regression models were developed for each individual cut, including fabrication byproduct items (bone, fat trim) to estimate the percentage yield of those cuts reported by USDA Market News. Strip loin, top sirloin butt, and gooseneck rounds from heifers tended to have a higher percentage yield at the same USDA yield grade than the same cuts from steers, possibly resulting from increased fat deposition on heifers. Percentage of fat trimmed from dairy steers was 2 to 3% lower than that from other sex-class/carcass types; however, due to increased percentage of bone and less muscle, dairy steers were lower-yielding. Fat trimmed from carcasses ranged from 7.9 to 15.6% as the maximum trim level decreased from 2.54 to .64 cm.     

Using current on-line carcass evaluation parameters to estimate boneless and bone-in pork carcass yield as influenced by trim level.

The objective of this study was to develop prediction equations for estimating proportional carcass yield to a variety of external trim levels and bone-in and boneless pork primal cuts. Two hundred pork carcasses were selected from six U.S. pork processing plants and represented USDA carcass grades (25% USDA #1, 36% USDA #2, 25% USDA #3, and 14% USDA #4). Carcasses were measured (prerigor and after a 24 h chill) for fat and muscle depth at the last rib (LR) and between the third and fourth from last rib (TH) with a Hennessy optical grading probe (OGP). Carcasses were shipped to Texas A&M University, where one was randomly assigned for fabrication. Selected sides were fabricated to four lean cuts (ham, loin, Boston butt, and picnic shoulder) then fabricated progressively into bone-in (BI) and boneless (BL) four lean cuts (FLC) trimmed to .64, .32, and 0 cm of s.c. fat, and BL 0 cm trim, seam fat removed, four lean cuts (BLS-OFLC). Total dissected carcass lean was used to calculate the percentage of total carcass lean (PLEAN). Lean tissue subsamples were collected for chemical fat-free analysis and percentage carcass fat-free lean (FFLEAN) was determined. Longissimus muscle area and fat depth also were collected at the 10th and 11th rib interface during fabrication. Regression equations were developed from linear carcass and OGP measurements predicting FLC of each fabrication point. Loin muscle and fat depths from the OPG obtained on warm, prerigor carcasses at the TH interface were more accurate predictors of fabrication end points than warm carcass probe depth obtained at the last rib or either of the chilled carcass probe sites (probed at TH or LR). Fat and loin muscle depth obtained via OGP explained 46.7, 52.6, and 57.1% (residual mean square error [RMSE] = 3.30, 3.19, and 3.04%) of the variation in the percentage of BI-FLC trimmed to .64, .32, and 0 cm of s.c. fat, respectively, and 49.0, 53.9, and 60.7% (RMSE = 2.91, 2.81, and 2.69%) of the variation in the percentage of BL-FLC trimmed to .64, .32, and 0 cm of s.c. fat, respectively. Fat and loin muscle depth from warm carcass OGP probes at the TH interface accounted for 62.4 and 63.5% (RMSE = 3.38 and 3.27%) of the variation in PLEAN and FFLEAN, respectively. These equations provide an opportunity to estimate pork carcass yield for a variety of procurement end point equations using existing on-line techniques.     

Carcass traits, cut yields, and compositional end points in high-lean-yielding pork carcasses: effects of 10th rib backfat and loin eye area

Pork carcasses (n = 133) were used to investigate the influence of carcass fatness and muscling on composition and yields of pork primal and subprimal cuts fabricated to varying levels of s.c. fat. Carcasses were selected from commercial packing plants in the southeastern United States, using a 3 x 3 factorial arrangement with three levels of 10th rib backfat depth (< 2.03, 2.03 to 2.54, and > 2.54 cm) and three levels of loin eye area (LEA; < 35.5, 35.5 to 41.9, and > 41.9 cm2). Sides from the selected carcasses were shipped to the University of Georgia for carcass data collection by trained USDA-AMS and University of Georgia personnel and fabrication. Sides were fabricated to four lean cuts (picnic shoulder, Boston butt, loin, and ham) and the skinned belly. The four lean cuts were further fabricated into boneless cuts with s.c. fat trim levels of 0.64, 0.32, and 0 cm. The percentages of four lean cuts, boneless cuts (four lean cuts plus skinned, trimmed belly) at 0.64, 0.32, and 0 cm s.c. fat, fat-free lean, and total fat were calculated. Data were analyzed using a least squares fixed effects model, with the main effects of 10th rib backfat and LEA and their interaction. Fatness and muscling traits increased (P < 0.05) as 10th rib backfat and LEA category increased, respectively. However, fat depth measures were not affected greatly by LEA category, nor were muscling measures greatly affected by backfat category. The percentage yield of cuts decreased (P < 0.05) as backfat category increased. Cut yields from the picnic shoulder, Boston butt, and belly were not affected (P > 0.05) by LEA category, whereas the yield of boneless loin and ham increased (P < 0.05) as LEA category increased. Compositionally, the percentage of four lean cuts, boneless cuts at varying trim levels, and fat-free lean decreased incrementally (P < 0.05) as backfat depth increased, whereas parentage total fat and USDA grade increased (P < 0.05) as backfat depth increased. As LEA increased, percentage boneless cuts trimmed to 0.32 and 0 cm s.c. fat and fat-free lean increased and total fat decreased; however, the difference was only significant in the smallest LEA category. Collectively, these data show that decreased carcass fatness plays a greater role in increasing primal and subprimal cut yields and carcass composition than muscling even in lean, heavily muscled carcasses.     

Beef carcass composition of slaughter cattle differing in frame size, muscle score, and external fatness.

Commercial slaughter steers (n = 329) and heifers (n = 335) were selected to vary in slaughter frame size and muscle thickness score, as well as carcass adjusted 12th-rib fat thickness. After collection of USDA carcass grade data, one side of each carcass was fabricated into boneless primals, subprimals, and minor tissue components. Cuts were trimmed to 2.54, 1.27, and .64 cm of external fat, except for the knuckle, tri-tip, and tenderloin, which were trimmed of all fat. Forced four-variable regression equations were used to predict the percentage (chilled carcass weight basis) yield of boneless subprimals at the three fat trim levels as influenced by sex class, frame size, muscle score, and adjusted 12th-rib fat thickness. Independent variables that had the most influence on percentage yield of primals and boneless subprimals were adjusted 12th-rib fat thickness and sex class. Within the same phenotypic group, percentage of trimmable fat increased by 2.32% as 12th-rib fat thickness increased by .75 cm. Estimated percentage yield of the major subprimals from the loin and round tended to be higher or relatively equal for heifer carcasses at all trim levels compared with those subprimals from steer carcasses. Holding frame size, sex class, and fat thickness constant, there was a higher percentage yield of chuck roll, rib eye roll, and strip loin for carcasses from thick-muscled cattle than for those from average- and thin-muscled cattle. Frame size had little effect on percentage yield of boneless subprimals.     

An evaluation of the lamb vision system as a predictor of lamb carcass red meat yield percentage

An objective method for predicting red meat yield in lamb carcasses is needed to accurately assess true carcass value. This study was performed to evaluate the ability of the lamb vision system (LVS; Research Management Systems USA, Fort Collins, CO) to predict fabrication yields of lamb carcasses. Lamb carcasses (n = 246) were evaluated using LVS and hot carcass weight (HCW), as well as by USDA expert and on-line graders, before fabrication of carcass sides to either bone-in or boneless cuts. On-line whole number, expert whole-number, and expert nearest-tenth USDA yield grades and LVS + HCW estimates accounted for 53, 52, 58, and 60%, respectively, of the observed variability in boneless, saleable meat yields, and accounted for 56, 57, 62, and 62%, respectively, of the variation in bone-in, saleable meat yields. The LVS + HCW system predicted 77, 65, 70, and 87% of the variation in weights of boneless shoulders, racks, loins, and legs, respectively, and 85, 72, 75, and 86% of the variation in weights of bone-in shoulders, racks, loins, and legs, respectively. Addition of longissimus muscle area (REA), adjusted fat thickness (AFT), or both REA and AFT to LVS + HCW models resulted in improved prediction of boneless saleable meat yields by 5, 3, and 5 percentage points, respectively. Bone-in, saleable meat yield estimations were improved in predictive accuracy by 7.7, 6.6, and 10.1 percentage points, and in precision, when REA alone, AFT alone, or both REA and AFT, respectively, were added to the LVS + HCW output models. Use of LVS + HCW to predict boneless red meat yields of lamb carcasses was more accurate than use of current on-line whole-number, expert whole-number, or expert nearest-tenth USDA yield grades. Thus, LVS + HCW output, when used alone or in combination with AFT and/or REA, improved on-line estimation of boneless cut yields from lamb carcasses. The ability of LVS + HCW to predict yields of wholesale cuts suggests that LVS could be used as an objective means for pricing carcasses in a value-based marketing system.     

Influence of body condition score on carcass characteristics and subprimal yield from cull beef cows.

Mature beef cows (n = 83) were slaughtered to measure the influence of body condition score (BCS) on carcass characteristics and subprimal yields. All cows were weighed and assigned BCS, based on a 9-point scale, 24 h before slaughter. Cows were slaughtered, and, after a 48-h chilling period, quality and yield grade data were collected on the left side of each carcass. The right side was quartered, fabricated into primal cuts, and weighed. Each primal cut was further processed into boneless subprimal cuts, minor cuts, lean trim, fat, and bone. Cuts were progressively trimmed to 6.4 and 0 mm of external and visible seam fat. Weights were recorded at all stages of fabrication, and subprimal yields were calculated as a percentage of the chilled carcass weight. Live weight, carcass weight, dressing percentage, fat thickness, longissimus muscle area, muscle:bone ratio, and numerical yield grade increased linearly (P = .0001) and predicted cutability and actual muscle-to-fat ratio decreased linearly (P = .0001) as BCS increased from 2 to 8. Carcasses from BCS-8 cows had the most (P<.05) marbling. The percentage of carcasses grading U.S. Utility, or higher, was 16.7, 20.0, 63.6, 43.3, 73.3, 100.0, and 100.0% for cows assigned a BCS of 2, 3, 4, 5, 6, 7, and 8, respectively. At 6.4 mm of fat trim, carcasses from BCS-5 cows had higher (P<.05) shoulder clod yields than carcasses from cows having a BCS of 6, 7, and 8. Carcasses of BCS-2 cows had lower (P<.05) strip loin yields than carcasses from BCS-3, 4, 5, 6, and 7 cows. Top sirloin butt yields were higher (P<.05) for carcasses of BCS-2, 3, 4, and 5 cows than those of BCS-6, 7, or 8 cows. Carcasses from BCS-7 and 8 cows had lower (P<.05) tenderloin and inside round yields than carcasses of BCS-5, or less, cows. At both fat-trim levels, carcasses from BCS-5 cows had higher (P<.05) eye of round yields than cows assigned BCS of 2, 7, or 8. When subprimal cuts were trimmed to 6.4 mm of visible fat, carcasses from BCS-5 cows had higher (P<.05) total lean product yields than cows assigned a BCS of 2, 4, 7, and 8. Regardless of fat trim, total fat yields increased (P = .0001) and total bone yields decreased (P = .0001) linearly as BCS increased from 2 to 8. Although carcasses from BCS-5 and 6 cows had the highest yields of lean product, cattle producers and packers may benefit most by marketing and(or) purchasing BCS-6 cows because a higher percentage of their carcasses had quality characteristics deemed desirable for fabrication into boneless subprimal cuts.     

Development and validation of equations utilizing lamb vision system output to predict lamb carcass fabrication yields

This study was performed to validate previous equations and to develop and evaluate new regression equations for predicting lamb carcass fabrication yields using outputs from a lamb vision system-hot carcass component (LVS-HCC) and the lamb vision system-chilled carcass LM imaging component (LVS-CCC). Lamb carcasses (n = 149) were selected after slaughter, imaged hot using the LVS-HCC, and chilled for 24 to 48 h at -3 to 1 degrees C. Chilled carcasses yield grades (YG) were assigned on-line by USDA graders and by expert USDA grading supervisors with unlimited time and access to the carcasses. Before fabrication, carcasses were ribbed between the 12th and 13th ribs and imaged using the LVS-CCC. Carcasses were fabricated into bone-in subprimal/primal cuts. Yields calculated included 1) saleable meat yield (SMY); 2) subprimal yield (SPY); and 3) fat yield (FY). On-line (whole-number) USDA YG accounted for 59, 58, and 64%; expert (whole-number) USDA YG explained 59, 59, and 65%; and expert (nearest-tenth) USDA YG accounted for 60, 60, and 67% of the observed variation in SMY, SPY, and FY, respectively. The best prediction equation developed in this trial using LVS-HCC output and hot carcass weight as independent variables explained 68, 62, and 74% of the variation in SMY, SPY, and FY, respectively. Addition of output from LVS-CCC improved predictive accuracy of the equations; the combined output equations explained 72 and 66% of the variability in SMY and SPY, respectively. Accuracy and repeatability of measurement of LM area made with the LVS-CCC also was assessed, and results suggested that use of LVS-CCC provided reasonably accurate (R2 = 0.59) and highly repeatable (repeatability = 0.98) measurements of LM area. Compared with USDA YG, use of the dual-component lamb vision system to predict cut yields of lamb carcasses improved accuracy and precision, suggesting that this system could have an application as an objective means for pricing carcasses in a value-based marketing system.     

Retail cut yields of Rambouillet wether lambs fed the beta-adrenergic agonist L644,969.

Twenty Rambouillet wether lambs were given ad libitum access to a diet with (BAA, n = 10) or without (control, n = 10) 1 ppm of the beta-adrenergic agonist L644,969. Lambs were fed to a constant slaughter weight end point of 54.5 kg. Carcasses were fabricated to yield bone-in and boneless cuts that were trimmed progressively to 1.27, .64, .32, and .00 cm of s.c. fat remaining. Addition of BAA did not affect growth traits. Actual and adjusted fat thickness, body wall thickness, and percentage of kidney-pelvic fat did not differ between control and BAA lambs. However, BAA increased longissimus muscle area, longissimus muscle depth, and leg score while decreasing USDA yield grade. The BAA increased carcass conformation scores and decreased flank lean color scores. No other carcass quality measurements were affected by BAA. Addition of BAA did not affect overall carcass yields of bone-in retail cuts. However, BAA increased overall carcass yields of boneless retail cuts regardless of fat trim level. The BAA increased bone-in leg yield. Yield of boneless sirloin, bone-in loin and boneless loin were not affected by BAA. For these cuts, the percentage change from the control was highly dependent on fat trim level. There was no difference in short-cut, shank-off, semiboneless leg yield between control and BAA. Addition of BAA did not affect yield of bone-in rack regardless of fat trim level. However, BAA greatly increased yield of boneless ribeye. The BAA did not affect yield of bone-in or boneless shoulder.(ABSTRACT TRUNCATED AT 250 WORDS)     

Gluteus medius and rump fat depths as additional live animal ultrasound measurements for predicting retail product and trimmable fat in beef carcasses.

This study was conducted to determine the ability of additional ultrasound measures to enhance the prediction accuracy of retail product and trimmable fat yields based on weight and percentage. Thirty-two Hereford-sired steers were ultrasonically measured for 12th-rib fat thickness, longissimus muscle area, rump fat thickness, and gluteus medius depth immediately before slaughter. Chilled carcasses were evaluated for USDA yield grade factors and then fabricated into closely trimmed, boneless subprimals with 0.32 cm s.c. fat. The kilogram weight of end-point product included the weight of trimmed, boneless subprimals plus lean trim weights, chemically adjusted to 20% fat, whereas the fat included the weight of trimmed fat plus the weight of fat in the lean trim. Prediction equations for carcass yield end points were developed using live animal or carcass measurements, and live animal equations were developed including ultrasound ribeye area or using only linear measurements. Multiple regression equations, with and without ultrasound rump fat thickness and gluteus medius depth, had similar R2 values when predicting kilograms of product and percentages of product, suggesting that these alternative variables explained little additional variation. Final unshrunk weight and ultrasound 12th-rib fat thickness explained most of the variation when predicting kilograms of fat. Rump fat and gluteus medius depth accounted for an additional 10% of the variation in kilograms of fat, compared with the equation containing final weight, ultrasound ribeye area, and ultrasound 12th-rib fat thickness; however, the two equations were not significantly different. Prediction equations for the cutability end points had similar R2 values whether live animal ultrasound measurements or actual carcass measurements were used. However, when ultrasound ribeye area was excluded from live animal predictions, lower R2 values were obtained for kilograms of product (0.81 vs 0.67) and percentages of product (0.41 vs 0.17). Conversely, the exclusion of ultrasound ribeye area had little effect on the prediction accuracy for kilograms of fat (0.75 vs 0.74) and percentage fat (0.50 vs 0.40). These data substantiate the ability of live animal ultrasound measures to accurately assess beef carcass composition and suggest that the alternative ultrasound measures, rump fat and gluteus medius depth, improve the accuracy of predicting fat-based carcass yields.     

Dual-component video image analysis system (VIASCAN) as a predictor of beef carcass red meat yield percentage and for augmenting application of USDA yield grades

An improved ability to quantify differences in the fabrication yields of beef carcasses would facilitate the application of value-based marketing. This study was conducted to evaluate the ability of the Dual-Component Australian VIASCAN to 1) predict fabricated beef subprimal yields as a percentage of carcass weight at each of three fat-trim levels and 2) augment USDA yield grading, thereby improving accuracy of grade placement. Steer and heifer carcasses (n = 240) were evaluated using VIASCAN, as well as by USDA expert and online graders, before fabrication of carcasses to each of three fat-trim levels. Expert yield grade (YG), online YG, VIASCAN estimates, and VIASCAN estimated ribeye area used to augment actual and expert grader estimates of the remaining YG factors (adjusted fat thickness, percentage of kidney-pelvic-heart fat, and hot carcass weight), respectively, 1) accounted for 51, 37, 46, and 55% of the variation in fabricated yields of commodity-trimmed subprimals, 2) accounted for 74, 54, 66, and 75% of the variation in fabricated yields of closely trimmed subprimals, and 3) accounted for 74, 54, 71, and 75% of the variation in fabricated yields of very closely trimmed subprimals. The VIASCAN system predicted fabrication yields more accurately than current online yield grading and, when certain VIASCAN-measured traits were combined with some USDA yield grade factors in an augmentation system, the accuracy of cutability prediction was improved, at packing plant line speeds, to a level matching that of expert graders applying grades at a comfortable rate.     

Predicting carcass composition of beef cattle using ultrasound technology.

We evaluated 20 slaughtered cattle with ultrasound before hide removal to predict fat thickness and ribeye area at the 12th rib for possible use in carcass composition prediction. Carcasses were fabricated into boneless subprimals that were trimmed progressively from 2.54 to 1.27 to .64 cm maximum fat trim levels. Stepwise regression was used to indicate the relative importance of variables in a model designed to estimate the percentage of boneless subprimals from the carcass at different external fat trim levels. Variables included those obtained on the slaughter floor (ultrasound fat thickness and ribeye area; estimated percentage of kidney, pelvic, and heart [KPH] fat; and warm carcass weight) and those obtained from carcasses following 24 h in the chill cooler (actual fat thickness, actual ribeye area, estimated percentage of KPH fat, warm carcass weight, and marbling score). At all different subprimal trim levels, percentage KPH was the first variable to enter the model. In the models using measures taken on the slaughter floor, ultrasound fat thickness was the only other variable to enter the model. Ultrasound fat thickness increased R2 and decreased residual standard deviation (RSD) in models predicting subprimals at 2.54-cm maximum fat trim; however, at 1.27- and .64-cm trim levels, R2 and RSD increased. Models using the same two variables (except actual fat instead of ultrasound) in the cooler were similar to those using data from the slaughter floor. However, as more cooler measurement variables entered the models, R2 increased and RSD decreased, explaining a greater amount of the variation in the equation. Ultrasonic evaluation on the slaughter floor may be of limited application compared with the greater accuracy found in chilled carcass assessment.     

Prediction of retail product and trimmable fat yields from the four primal cuts in beef cattle using ultrasound or carcass data

The most widely used system to predict percentage of retail product from the four primal cuts of beef is USDA yield grade. The purpose of this study was to determine whether routine ultrasound measurements and additional rump measurements could be used in place of the carcass measurements used in the USDA yield grade equation to more accurately predict the percentage of saleable product from the four primals. This study used market cattle (n = 466) consisting of Angus bulls, Angus steers, and crossbred steers. Live animal ultrasound measures collected within 7 d of slaughter were 1) scan weight (SCANWT); 2) 12th- to 13th-rib s.c. fat thickness (UFAT); 3) 12th- to 13th-rib LM area (ULMA); 4) s.c. fat thickness over the termination of the biceps femoris in the rump (URFAT; reference point); 5) depth of gluteus medius under the reference point (URDEPTH); and 6) area of gluteus medius anterior to the reference point (URAREA). Traditional carcass measures collected included 1) HCW; 2) 12th-to 13th-rib s.c. fat thickness (CFAT); 3) 12th- to 13th-rib LM area (CLMA); and 4) estimated percentage of kidney, pelvic, and heart fat (CKPH). Right sides of carcasses were fabricated into subprimal cuts, lean trim, fat, and bone. Weights of each component were recorded, and percentage of retail product from the four primals was expressed as a percentage of side weight. A stepwise regression was performed using data from cattle (n = 328) to develop models to predict percentage of retail product from the four primals based on carcass measures or ultrasound measures, and comparisons were made between the models. The most accurate carcass prediction equation included CFAT, CKPH, and CLMA (R2 = 0.308), whereas the most accurate live prediction equation included UFAT, ULMA, SCANWT, and URAREA (R2 = 0.454). When these equations were applied to a validation set of cattle (n = 138), the carcass equation showed R2 = 0.350, whereas the ultrasound data showed R2 = 0.460. Ultrasound measures in the live animal were potentially more accurate predictors of retail product than measures collected on the carcass.     

Real-time augmentation of USDA yield grade application to beef carcasses using video image analysis

In two phases, this study assessed the ability of two video image analysis (VIA) instruments, VIASCAN and Computer Vision System (CVS), to augment assignment of yield grades (YG) to beef carcasses to 0.1 of a YG at commercial packing plant speeds and to test cutout prediction accuracy of a YG augmentation system that used a prototype augmentation touchpanel grading display (designed to operate commercially in real-time). In Phase I, beef carcasses (n = 505) were circulated twice at commercial chain speeds (340 carcasses per hour) by 12 on-line USDA graders. During the first pass, on-line graders assigned a whole-number YG and a quality grade (QG) to carcasses as they would normally. During the second pass, on-line graders assigned only adjusted preliminary yield grades (APYG) and QG to carcasses, whereas the two VIA instruments measured the longissimus muscle area (LMA) of each carcass. Kidney, pelvic, and heart fat (KPH) was removed and weighed to allow computation of actual KPH percentage. Those traits were compared to the expert YG and expert YG factors. On-line USDA graders' APYG were closely related (r = 0.83) to expert APYG. Instrument-measured LMA were closely related (r = 0.88 and 0.94; mean absolute error = 0.3 and 0.2 YG units, for VIASCAN and CVS, respectively) to expert LMA. When YG were augmented using instrument-measured LMA and computed either including or neglecting actual KPH percentage, YG were closely related (r = 0.93 and 0.92, mean absolute error = 0.32 and 0.40 YG units, respectively, using VIASCAN-measured LMA; r = 0.95 and 0.94, mean absolute error = 0.24 and 0.34 YG units, respectively, using CVS-measured LMA) to expert YG. In Phase II, augmented YG were assigned (0.1 of a YG) to beef carcasses (n = 290) at commercial chain speeds using VIASCAN and CVS to determine LMA, whereas APYG and QG were determined by online graders via a touch-panel display. On-line grader YG (whole-number), expert grader YG (to the nearest 0.1 of a YG), and VIASCAN- and CVS-augmented YG (to the nearest 0.1 of a YG) accounted for 55, 71, 60, and 63% of the variation in fabricated yields of closely trimmed subprimals, respectively, suggesting that VIA systems can operate at current plant speeds and effectively augment official USDA application of YG to beef carcasses.     

Using live estimates and ultrasound measurements to predict beef carcass cutability

Commercial slaughter steers (n = 329) and heifers (n = 335) were selected to vary in frame size, muscle score, and carcass fat thickness to study the effectiveness of live evaluation and ultrasound as predictors of carcass composition. Three trained personnel evaluated cattle for frame size, muscle score, fat thickness, longissimus muscle area, and USDA quality and yield grade. Live and carcass real-time ultrasound measures for 12th-rib fat thickness and longissimus muscle area were taken on a subset of the cattle. At the time of slaughter, carcass ultrasound measures were taken at "chain speed." After USDA grade data were collected, one side of each carcass was fabricated into boneless primals/subprimals and trimmed to .64 cm of external fat. Simple correlation coefficients showed a moderately high positive relationship between 12th rib fat thickness and fat thickness measures obtained from live estimates (r = .70), live ultrasound (r = .81), and carcass ultrasound (r = .73). The association between estimates of longissimus muscle area and carcass longissimus muscle area were significant (P < .001) and were higher for live evaluation (r = .71) than for the ultrasonic measures (live ultrasound, r = .61; carcass ultrasound, r = .55). Three-variable regression equations, developed from the live ultrasound measures, explained 57% of the variation in percentage yield of boneless subprimals, followed by live estimates (R2 = .49) and carcass ultrasound (R2 = .31). Four-variable equations using frame size, muscle score, and selected fat thickness and weight measures explained from 43% to 66% of the variation for the percentage yield of boneless subprimals trimmed to .64 cm. Live ultrasound and(or) live estimates are viable options for assessing carcass composition before slaughter.     

Yield grades and cutability of carcass beef with and without kidney and pelvic fat

Cut-out data from 2,550 steer carcasses representing British, Continental and Zebu breeding were analyzed to evaluate yield grade classes based on three equations: 1) Ya = 2.5 + .984 AFT + .0084 HCW - .05 REA + .2 KPF; 2) Yb = same as Ya with intercept changed to 3.2 and KPF deleted; 3) Yc = 3.0 + .984 AFT + .0041 HCW - .03 REA, as proposed by USDA in 1984; where AFT = adjusted fat thickness (cm), HCW = hot carcass weight (kg), REA = rib-eye area (cm2), and KPF = kidney and pelvic fat (%). Essentially boneless, closely trimmed (8 mm) roasts and steaks of the four major retail cuts (MRC) were made from one side of each carcass. Cutability was calculated as: Ca, % = 100 (MRC/side) and Cb,c % = 100 (MRC/side, KPF removed). Cutability increased (P less than .01) an average of two percentage points when KPF was removed. In general, removing KPF from the estimation of cutability and changing the coefficients for REA and HCW resulted in a decrease in the number of carcasses yield graded 1 or 4 and an increase in the number of carcasses in yield grade 2. Redistribution of carcasses was greater for Yc than for Yb. Carcasses classified with equation Yc tended (P greater than .05) to have greater cutability in yield grades 1 and 2, and lower (P less than .01) cutability in yield grades 3, 4 and 5.(ABSTRACT TRUNCATED AT 250 WORDS)     

Characterization of biological types of cattle (Cycle VI): carcass, yield, and longissimus palatability traits

Carcass (n = 568) and longissimus thoracis palatability (n = 460) traits from F1 steers obtained from mating Hereford (H), Angus (A), and U.S. Meat Animal Research Center (MARC) III cows to H, A, Norwegian Red (NR), Swedish Red and White (RW), Friesian (F), or Wagyu (W) sires were compared. Data were adjusted to constant age (471 d), carcass weight (356 kg), fat thickness (1.0 cm), percentage of fat trim (24%), and marbling (Small35) end points. For Warner-Bratzler shear force and trained sensory panel traits, data were obtained on longissimus thoracis steaks stored at 2 degrees C for 14 d postmortem. The following comparisons were from the age-constant end point. Carcasses from H- and A-sired steers (377 and 374 kg, respectively) were the heaviest (P < 0.05) and carcasses from W-sired steers (334 kg) were the lightest (P < 0.05). A greater (P < 0.05) percentage of carcasses from A- and W-sired steers graded USDA Choice (88 and 85%, respectively) than carcasses from other sire breeds (52 to 71%). Adjusted fat thickness for carcasses from A-sired steers (1.3 cm) was highest (P < 0.05), followed by H-sired steers (1.1 cm) and W- and F-sired steers (0.9 cm); NR- and RW-sired steers (0.8 cm) had the lowest (P < 0.05) adjusted fat thickness. Longissimus thoracis area was not different (P > 0.05) among sire breeds (mean = 80.6 cm2). Carcass yield of boneless, totally trimmed retail product was least (P < 0.05) for A-sired steers (60.1%), intermediate for H-sired steers (61.5%), and similar (P > 0.05) for all other sire breeds (62.5 to 62.8%). Longissimus thoracis steaks from carcasses of A- (3.7 kg) and W-sired (3.7 kg) steers had lower (P < 0.05) shear force values than longissimus thoracis steaks from other sire breeds (4.1 to 4.2 kg). Trained sensory panel tenderness, juiciness, or beef flavor intensity ratings for longissimus thoracis steaks did not differ (P > 0.05) among the sire breeds. Sire breed comparisons were affected by adjusting data to other end points. Heritability estimates for various carcass, yield, and palatability traits ranged from very low (h2 = 0.06 for percentage of kidney, pelvic, and heart fat) to relatively high (h2 = 0.71 for percentage of retail product yield). Relative to the other sire breeds, W-sired steers had the highest percentage of USDA Choice, Yield grade 1 and 2 carcasses, but their carcasses were the lightest.