Feasibility of Attenuation Coefficient and Envelope Signal-to-Noise Ratio Estimation for Tissue Characterization of Liver Steatosis

Abstract

Nonalcoholic fatty liver disease (NAFLD) has become the most common cause of chronic liver disease worldwide. In addition, conducting large-scale screening studies is still challenging due to the absence of reliable and affordable noninvasive tools for NAFLD staging. Conventional ultrasound is commonly the preferred first imaging modality used to evaluate liver steatosis. However, it is mainly a qualitative technique that depends on operator's skill and experience, resulting in variable results and reproducibility. In this study, we assess the application of attenuation coefficient slope (ACS) estimation by the regularized spectral log difference (RSLD) technique and computation of the envelope signal-to-noise ratio (SNR) for tissue characterization of NAFLD in a murine animal model. This approach was tested using rats that were fed either a control or methionine and choline deficient (MCD) diet. Ultrasound measurements were then performed at 0 (baseline), 2, and 6 weeks. After the final imaging session, animals were humanely euthanized and livers excised for histological processing. Results showed that ACS and SNR increased by approximately 0.2 dB/cm/MHz and 0.2 a.u., respectively, for each 25% increment in fat fraction percentage in diet group over the weeks. Moreover, control and diet groups were statistically different at weeks 2 $(p < 0.05)$ and 6 $(p < 0.01)$ for both parameters. Overall, ACS and SNR were found to be encour-aging for determining relative contents of liver fat and noninvasive grading of steatosis.