Changes in Ultrasound Backscattering Statistics at Different Degrees of Hepatic Steatosis

• Main Authors: Jui Fang, 方銳
• Other Authors: P. H. Tsui
• Format: Others
• Language: en_US
• Published: 2018
• Online Access: http://ndltd.ncl.edu.tw/handle/eubbg2

博士 === 長庚大學 === 生物醫學工程博士學位學程 === 106 === Hepatic steatosis is an abnormal state, where excessive lipid mass accumulates in hepatocyte droplets. Backscattered ultrasound signals received from the liver contain information regarding the degree of steatosis in the liver. Homodyned-K (HK) and its approximation, Nakagami distribution, have been used as general models for ultrasound scattering. Their parameters—the μ and m parameters for HK and Nakagami distributions, respectively—are sensitive to scatterer concentrations. However, only the μ parameter effectively differentiates between different media with high scatterer concentrations, whereas the m parameter becomes saturated under fully developed speckle. This study explored the diagnostic performance of μ parametric imaging in the assessment of hepatic steatosis. The μ parameter was estimated using two methods: the simplicity method, based on the first three integer moments (denoted as μ_FTM), and the precision method based on the first moment of the intensity and two log-moments (denoted as μ_XU). Nakagami m parametric images were also constructed for comparison. Phantoms, containing 0.13, 0.52, 1, 2, 3, 4, 5, and 6 g of graphite powder, were initially conducted using the sliding window technique to determine an appropriate window size length (WSL) for HK parametric imaging. Subsequently, hepatic steatosis was induced in male Wistar rats fed a methionine-and-choline-deficient (MCD) diet in two schedules. The first schedule for different steatosis stages comprised six groups fed MCD for 0 (i.e., normal control), 1, 2, 4, 6, and 8 weeks (n = 36; six rats in each group). Second schedule for severe steatosis (n = 30; 6 in each group) comprised six groups fed MCD for 4, 8, 12, 16, 20, and 24 weeks. Area under the receiver operating characteristic curve (AUC) was used to assess diagnostic performance levels of μ parametric imaging. Finally, the effects of a novel intervention, whole body vibration (WBV), on hepatic steatosis was compared with traditional swim exercise through μ (μ_XU and μ_FTM) parametric imaging and corresponding parameters. Obesity-related hepatic steatosis was induced in 54 male C57BL/6 mice fed a high-fat diet for 4, 8, and 16 weeks. Three groups, sedentary, swim, and WBV, were used (18 mice in each group). Three parameters, μ_XU, μ_FTM, and m, were correlated with the histological features of the droplets (size, area ratio, and number). The phantom results showed that appropriate WSLs for μ_XU and μ_FTM parametric imaging were five and seven times the pulse length of the transducer, respectively. The animal experiments demonstrated promising diagnostic performances of μ_XU, μ_FTM, and m parametric imaging (AUCs > 0.8) for staging steatosis in rats. However, high linearity was observed only between μ_XU and droplet size (r = 0.728). The WBV experiments revealed that WBV and swimming demonstrated significant effects on hepatic steatosis in at least 16 weeks. Moreover, some results were only found in μ_XU. The findings suggest that ultrasound HK μ parametric imaging based on the XU method has a great potential for clinical diagnosis of mild to severe hepatic steatosis.