Abstract

Twenty-one Dorset Down ewe hoggets (49.4kg body weight, 9months old) were scanned by ultrasound and X-ray computer tomography (CT). Total carcass tissue (fat, muscle and bone) weights were assessed by the Cavalieri procedure from 16-18 CT slices per animal. Additional CT slices were taken at four anatomical reference sites; 7th thoracic (TV7), 2nd lumbar and 5th lumbar vertebrae (LV5), and mid-shaft of the femur. Subcutaneous fat depth and muscle depth were assessed over the last rib using ultrasound. Body weight, ultrasound measurements and reference slice carcass tissue areas were used to predict total carcass tissue weights by multiple regression techniques. In combination with BW, single CT slices predicted tissue weight more accurately than ultrasound measurements with BW for fat (R2=72-84% vs. 63%) but not for muscle (R2=79-87% vs. 83%) or bone (R2=62-66% vs. 65%). The best combination of two CT slices was TV7 & LV5. In combination with BW these improved prediction accuracy further (fat, R2=92%, muscle, R2=91%, bone, R2=71%). It is possible that data from a single CT slice predict carcass tissue weight less accurately than expected due to anatomical registration errors despite CT producing better quality images than ultrasound. These errors are attributed to scanning protocol constraints for CT equipment. Two or more CT slices need to be scanned in order to predict tissue weight more accurately than ultrasound.

MJ, Young, SJ Nsoso, CM Logan, and PR Beatson

Proceedings of the New Zealand Society of Animal Production, Volume 56, , 205-211, 1996