Texture-based classification of liver fibrosis using MRI.

PURPOSE: To investigate the ability of texture analysis of MRI images to stage liver fibrosis. Current noninvasive approaches for detecting liver fibrosis have limitations and cannot yet routinely replace biopsy for diagnosing significant fibrosis. MATERIALS AND METHODS: Forty-nine patients with a range of liver diseases and biopsy-confirmed fibrosis were enrolled in the study. For texture analysis all patients were scanned with a T2 -weighted, high-resolution, spin echo sequence and Haralick texture features applied. The area under the receiver operating characteristics curve (AUROC) was used to assess the diagnostic performance of the texture analysis. RESULTS: The best mean AUROC achieved for separating mild from severe fibrosis was 0.81. The inclusion of age, liver fat and liver R2 variables into the generalized linear model improved AUROC values for all comparisons, with the F0 versus F1-4 comparison the highest (0.91). CONCLUSION: Our results suggest that a combination of MRI measures, that include selected texture features from T2 -weighted images, may be a useful tool for excluding fibrosis in patients with liver disease. However, texture analysis of MRI performs only modestly when applied to the classification of patients in the mild and intermediate fibrosis stages.

Stereological Analysis of Liver Biopsy Histology Sections as a Reference Standard for Validating Non-Invasive Liver Fat Fraction Measurements by MRI.

BACKGROUND AND AIMS: Validation of non-invasive methods of liver fat quantification requires a reference standard. However, using standard histopathology assessment of liver biopsies is problematical because of poor repeatability. We aimed to assess a stereological method of measuring volumetric liver fat fraction (VLFF) in liver biopsies and to use the method to validate a magnetic resonance imaging method for measurement of VLFF. METHODS: VLFFs were measured in 59 subjects (1) by three independent analysts using a stereological point counting technique combined with the Delesse principle on liver biopsy histological sections and (2) by three independent analysts using the HepaFat-Scan® technique on magnetic resonance images of the liver. Bland Altman statistics and intraclass correlation (IC) were used to assess the repeatability of each method and the bias between the methods of liver fat fraction measurement. RESULTS: Inter-analyst repeatability coefficients for the stereology and HepaFat-Scan® methods were 8.2 (95% CI 7.7-8.8)% and 2.4 (95% CI 2.2-2.5)% VLFF respectively. IC coefficients were 0.86 (95% CI 0.69-0.93) and 0.990 (95% CI 0.985-0.994) respectively. Small biases (≤3.4%) were observable between two pairs of analysts using stereology while no significant biases were observable between any of the three pairs of analysts using HepaFat-Scan®. A bias of 1.4±0.5% VLFF was observed between the HepaFat-Scan® method and the stereological method. CONCLUSIONS: Repeatability of the stereological method is superior to the previously reported performance of assessment of hepatic steatosis by histopathologists and is a suitable reference standard for validating non-invasive methods of measurement of VLFF.

Diagnostic performance of a rapid magnetic resonance imaging method of measuring hepatic steatosis.

OBJECTIVES: Hepatic steatosis is associated with an increased risk of developing serious liver disease and other clinical sequelae of the metabolic syndrome. However, visual estimates of steatosis from histological sections of biopsy samples are subjective and reliant on an invasive procedure with associated risks. The aim of this study was to test the ability of a rapid, routinely available, magnetic resonance imaging (MRI) method to diagnose clinically relevant grades of hepatic steatosis in a cohort of patients with diverse liver diseases. MATERIALS AND METHODS: Fifty-nine patients with a range of liver diseases underwent liver biopsy and MRI. Hepatic steatosis was quantified firstly using an opposed-phase, in-phase gradient echo, single breath-hold MRI methodology and secondly, using liver biopsy with visual estimation by a histopathologist and by computer-assisted morphometric image analysis. The area under the receiver operating characteristic (ROC) curve was used to assess the diagnostic performance of the MRI method against the biopsy observations. RESULTS: The MRI approach had high sensitivity and specificity at all hepatic steatosis thresholds. Areas under ROC curves were 0.962, 0.993, and 0.972 at thresholds of 5%, 33%, and 66% liver fat, respectively. MRI measurements were strongly associated with visual (r(2) = 0.83) and computer-assisted morphometric (r(2) = 0.84) estimates of hepatic steatosis from histological specimens. CONCLUSIONS: This MRI approach, using a conventional, rapid, gradient echo method, has high sensitivity and specificity for diagnosing liver fat at all grades of steatosis in a cohort with a range of liver diseases.