ABSTRACT. The usefulness of computed tomography (CT) for the diagnosis of feline hepatic lipidosis (FHL) was evaluated. Liver CT number was 54.7 ± 5.6 HU (mean ± SD) in 26 healthy cats. We fast 6 healthy cats for 72 hr to induced FHL experimentally and the cats were assessed by CT and serum biochemical analysis. Liver CT number of the six cats was 53.8 ± 3.0 HU before fasting, 46.8 ± 2.4 HU after fasting, and 50.2 ± 3.6 HU two weeks after restarted feeding. The decreased CT number was associated with the elevation of serum non-esterified fatty acid (NEFA) and β-hydroxybutyrate levels. These results indicate that measurement of CT number of the liver is an effective procedure for the diagnosis of FHL. KEY WORDS: computed tomography, feline hepatic lipidosis.J. Vet. Med. Sci. 67(11): 1163-1165, 2005 Feline hepatic lipidosis (FHL) is a metabolic disorder caused by excessive accumulation of triglycerides in hepatocytes that ultimately impairs liver function. Though, global surveys of feline hepatobiliary diseases have not been done, it is generally agreed that FHL is the most common liver disease in cats [7]. Diagnosis of FHL is based on cytology or histopathologic evaluation of liver specimen [5]. Hepatic biopsy can be performed by percutaneous fine-needle aspiration (FNA), but it lacks histopathological information, and regarded as least accurate diagnostic procedure for liver diseases [3,5]. In contrast, tissue biopsy with core needles, laparoscopy or laparotomy is more invasive and sometimes requires sedation or anesthesia. Moreover, histopathological estimation of fatty droplets accumulation in the hepatic tissue seems subjective.Fat is relatively radiolucent, so that objects rich in fat component are described as lower attenuation by Computed Tomography (CT) [4]. CT has been applied for evaluation of various liver diseases in human, including diagnosis of fatty liver [6]. In this study, we investigated the usefulness of CT for the diagnosis of FHL.Twenty-six clinical healthy cats and two cats with suspected lipidosis were investigated. Control cats were kept in our laboratory (n=6) or provided by the owner (n=20). These cats have normal plasma levels of alkaline phosphatase (ALP), alanine aminotransferase (ALT), asparate aminotransferase (AST) and bilirubin. CT scan was performed with multislice helical CT (Siemens Somatom 2, Siemens, Munich, Germany) at 130 kV, 60 mAs, and 5 mm slice thickness without contrast enhancement. Sedation with medetomidine-HCl (Domitor, Meiji seika, Tokyo,Japan) was performed if necessary. A transverse section image of liver was obtained at just caudal to the xiphoid process. Area of liver parenchyma was selected manually from the image to avoid blood vessels and area of inhomogeneity. The mean CT number and the standard deviation of the area were calculated with computer assisted image analyzer and expressed in Hounsfield unit (HU).CT number in control cats was 54.7 ± 5.6 HU (mean ± SD), ranged from 42.6 to 64.8 HU (Fig. 1) without significant difference between male and female cat...