Assessment of dairy sheep carcass composition with X-ray computed tomography

Abstract. The objective was to test a CT-acquisition protocol, with automated settings, for post-mortem assessment of dairy sheep carcasses using also two image analysis protocols (IAP). Twenty-six sheep of both sexes were used. After slaughter, hot carcass weight was recorded. Carcasses were chilled for 24h, then subjected to CT-scanning (Optima CT520, GE). The two CT-acquisition protocols (CT1, CT2) were tested on 4 carcasses. CT1 was automated and involved dose efficiency parameters, commonly used in medical practice (ranging field of view FoV and tube current-TC, standard tube tension-TT of 120 kV). In CT2, parameters were manually set as defaults throughout scanning (Large FoV, TC: 80 mA, TT: 140 kV). Slice thickness was 0.625 mm in both CT1 and CT2. The remaining 22 carcasses were scanned only with CT1. Two IAPs were tested on CT1 images of all carcasses; different thresholds were used to allocate voxels to muscle (IAP1: 0-120 HU, IAP2: 0-200 HU) and bone tissue (IAP1: HU>120, IAP2: HU>200), whereas fat tissue thresholds were standard; -200 to -1 HU. Tissue volumes and weights were calculated based on voxel volume, number of tissue voxels and tissue densities. Bone weight was calculated by subtracting fat and muscle weight from carcass weight. Paired sample t-tests and Wilcoxon signed-rank tests were used to assess estimation differences between CT1 and CT2. Mann-Whitney U tests were used for tissue estimations of both IAPs with available dissection data from similar sheep of the same population (n=101). Differences between CT1 and CT2 were not statistically significant. Bone estimations of both IAPs and muscle content estimation of IAP1 significantly differed from dissection data (P<0.01). Both CT-protocols are suitable for post-mortem assessment of dairy sheep carcass composition; hence automated CT1 can replace CT2. IAP2 better estimates muscle tissue compared to IAP1. However, bone tissue estimations are more complex. Work was funded by GreQuM project (T1EDK-05479), co-financed by Greece and EU (EPAnEK 2014-2020, Partnership Agreement 2014-2020).