Technical note: A novel technique to assess internal body fat of cattle by using real-time ultrasound.

The objectives of this study were to describe a system to assess KPH fat by using real-time ultrasound (RTU) and to develop equations to predict total physical separable internal fat (IFAT) based on ultrasound measurements. Data for this study were obtained from 24 Angus steers fed either hay- or corn-based diets during the backgrounding phase. Steers were serially slaughtered in 3 groups: at weaning (baseline), then at 4 and 8 mo after weaning. A fourth group was composed of 4 steers from the hay-fed group that were slaughtered at approximately 10 mo after weaning. The RTU measurements were collected every 2 mo, with a preslaughter scan approximately 7 d before the slaughter time. The RTU measurements consisted of 12th- to 13th-rib backfat thickness, 12th to 13th ribeye area, percentage of intramuscular fat, and kidney fat depth, which was measured in a cross-sectional image collected between the first lumbar vertebra and the 13th rib. For kidney fat, the ultrasound probe was placed on the flank region approximately 15 cm from the midline of the animal. Images were stored in the ultrasound console, and measurements were taken between the ventral part of the iliocostalis muscle and the end of the KPH fat at the chute side. The relationship between carcass and ultrasound measurements in the depths of kidney fat (cKFd and uKFd, respectively) had an r(2) of 0.93, with a root mean square error (RMSE) of 1.14 cm. An allometric regression between carcass KPH weight (cKPHwt) and cKFd was identified, and the untransformed regression had an r(2) of 0.96. The linear regression between total IFAT and cKPHwt had an r(2) of 0.97, with an RMSE of 2.67 kg. Therefore, a system was developed to predict IFAT from uKFd measurements by combining these equations. Additionally, a single linear regression between IFAT and uKFd measurements was developed (r(2) = 0.89, RMSE = 5.32 kg). Even though the system of equations had a lower RMSE of prediction and greater r(2) compared with the single linear regression (4.80 vs. 5.10 kg and 0.91 vs. 0.89, respectively), there was no difference between these methods in predicting IFAT (P = 0.4936) by using a pairwise mean square error of prediction analysis. Our results indicated that uKFd measurements can accurately and precisely predict the cKFd of steers consuming either high concentrate or forage rations. The results also showed that cKFd is highly correlated with cKPHwt, which can be used to estimate total IFAT. More research is needed to further evaluate this technique with different feeding strategies, breeds, and sexes.