Quantitative evaluation of hepatic steatosis using normalized local variance in a rat model: Comparison with histopathology as the reference standard

Abstract

Objective: To evaluate the diagnostic performance of the normalized local variance (NLV) ultrasound technique in the assessment of hepatic steatosis, and to identify the factors that influence the NLV value using histopathological examination as the reference standard. Materials and Methods: Forty male Sprague-Dawley rats were fed a methionine-choline-deficient diet for variable periods (0, 2, 4, 6, 8, 10, or 12 days or 2, 3, or 4 weeks; four rats per group). At the end of each diet duration, magnetic resonance spectroscopy (MRS) and NLV examination were performed. Thereafter, the rats were sacrificed and their livers were histopathologically evaluated. Receiver operating characteristic (ROC) curve analysis was performed to assess the diagnostic capability of the NLV value in the detection of varying degrees of hepatic steatosis. Univariate and multivariate linear regressions were used to determine the factors associated with the NLV value. Results: The areas under the ROC curve for the detection of mild, moderate, and severe hepatic steatosis were 0.953, 0.896, and 0.735, respectively. The NLV value showed comparable diagnostic performance to that of MRS in the detection of ≥ mild or ≥ moderate hepatic steatosis. Multivariate linear regression analysis revealed that the degree of hepatic steatosis was the only significant factor affecting the NLV value (p < 0.001). Conclusion: The NLV value of ultrasound demonstrated satisfactory diagnostic performance in the assessment of varying degrees of hepatic steatosis. The degree of hepatic steatosis was the only significant factor that affected the NLV value.