The repeatability and accuracy of ultrasound in measuring backfat of cattle.

Repeatability of ultrasound backfat measurements was determined on 217 cattle, and the association of ultrasound and carcass backfat measurements was evaluated with 580 animals. The correlation (repeatability) between consecutive ultrasound measurements was .975. Average difference between two ultrasound measures was .72 mm and error size was directly related (P less than .001) to the amount of backfat. Ultrasound measures averaged 8% (P less than .001) less than carcass measures. The average absolute difference between these two measures was 1.57 mm. Discrepancies were (P less than .001) larger when backfat was thicker; cattle with backfat measuring less than 10 mm averaged 1.43 mm absolute difference, whereas those with backfat greater than 10 mm averaged 1.89 mm. The variance of the discrepancies was partitioned into three components: ultrasound measurement error (14%), drift related to intercept and slope (15%), and residual (71%). The residual component may have represented errors in carcass measurement. In a subsequent trial with 175 cattle, average difference between ultrasound and carcass backfat measurements was reduced to 1.19 mm. Also, the slope and intercept virtually mimicked the isopach. Ultrasound measures can be more precise (minutely exact) and accurate (close to the true value) than carcass measures for assessing absolute backfat thickness in cattle. Elements of measurement protocol that may improve ultrasound technique are imparted in the manuscript.