¹College of Animal Science and Veterinary Medicine, Heilongjiang Bayi Agricultural University, Daqing, 163319, CHINA
²College of Animal Science and Veterinary Medicine, Jilin University, 5333 Xi’an Road, Changchun 130062, CHINA DOI : 10.9775/kvfd.2017.17932 Fatty liver disease is a common liver disease in humans and animals alike, and macrofauna are considered an ideal model of fatty liver disease for ultrasonography (US)-based diagnostic research. Here, a model of fatty liver disease in lactating thin-tail sheep induced by negative energy balance was developed by restricted feeding. Ultrasonographic images were divided into the following fat classes according to liver triglyceride concentrations (ratio of triglyceride weight to wet liver weight): healthy liver, <2%; mild hepatic steatosis (HS), ≥2% to 5%; moderate HS, ≥5% to 10%; and severe HS, ≥10%. Characteristics of Ultrasonographic images according to HS severity were evaluated. As the results, mild to severe HS was detected by restricted feeding. The sensitivity of 5 MHz for diagnostic differentiation of healthy liver, and mild, moderate, and severe HS were 66.7%, 60.0%, 66.7%, and 82.5%, respectively, with a moderate κ statistic ranging from 0.667 to 0.735 among observers. The sensitivity of 3.5 MHz for diagnosis of healthy liver, and mild, moderate, and severe HS were 69.7%, 60.5%, 81.5%, and 75.4%, respectively, with moderate κ statistics ranging from 0.575 to 0.684 among observers. The sensitivities of 3.5 and 5 MHz for moderate to severe HS reached 97.4% and 98.2%, respectively. In summary, visualization at 3.5 and 5 MHz was ideal for diagnosis of HS except mild cases with established US diagnostic standards. Furthermore, with the established model, newer methods, such as image digitizing analysis, can be used for US-based diagnosis of fatty liver in further investigations. Keywords : Ultrasound, Fatty liver, Sheep, Negative energy balance