Correlation of histologic, imaging, and artificial intelligence features in NAFLD patients, derived from Gd-EOB-DTPA-enhanced MRI: a proof-of-concept study.

OBJECTIVE: To compare unsupervised deep clustering (UDC) to fat fraction (FF) and relative liver enhancement (RLE) on Gd-EOB-DTPA-enhanced MRI to distinguish simple steatosis from non-alcoholic steatohepatitis (NASH), using histology as the gold standard. MATERIALS AND METHODS: A derivation group of 46 non-alcoholic fatty liver disease (NAFLD) patients underwent 3-T MRI. Histology assessed steatosis, inflammation, ballooning, and fibrosis. UDC was trained to group different texture patterns from MR data into 10 distinct clusters per sequence on unenhanced T1- and Gd-EOB-DTPA-enhanced T1-weighted hepatobiliary phase (T1-Gd-EOB-DTPA-HBP), then on T1 in- and opposed-phase images. RLE and FF were quantified on identical sequences. Differences of these parameters between NASH and simple steatosis were evaluated with χ2- and t-tests, respectively. Linear regression and Random Forest classifier were performed to identify associations between histological NAFLD features, RLE, FF, and UDC patterns, and then determine predictors able to distinguish simple steatosis from NASH. ROC curves assessed diagnostic performance of UDC, RLE, and FF. Finally, we tested these parameters on 30 validation cohorts. RESULTS: For the derivation group, UDC-derived features from unenhanced and T1-Gd-EOB-DTPA-HBP, plus from T1 in- and opposed-phase, distinguished NASH from simple steatosis (p ≤ 0.001 and p = 0.02, respectively) with 85% and 80% accuracy, respectively, while RLE and FF distinguished NASH from simple steatosis (p ≤ 0.001 and p = 0.004, respectively), with 83% and 78% accuracy, respectively. On multivariate regression analysis, RLE and FF correlated only with fibrosis (p = 0.040) and steatosis (p ≤ 0.001), respectively. Conversely, UDC features, using Random Forest classifier predictors, correlated with all histologic NAFLD components. The validation group confirmed these results for both approaches. CONCLUSION: UDC, RLE, and FF could independently separate NASH from simple steatosis. UDC may predict all histologic NAFLD components. CLINICAL RELEVANCE STATEMENT: Using gadoxetic acid-enhanced MR, fat fraction (FF > 5%) can diagnose NAFLD, and relative liver enhancement can distinguish NASH from simple steatosis. Adding AI may let us non-invasively estimate the histologic components, i.e., fat, ballooning, inflammation, and fibrosis, the latter the main prognosticator. KEY POINTS: • Unsupervised deep clustering (UDC) and MR-based parameters (FF and RLE) could independently distinguish simple steatosis from NASH in the derivation group. • On multivariate analysis, RLE could predict only fibrosis, and FF could predict only steatosis; however, UDC could predict all histologic NAFLD components in the derivation group. • The validation cohort confirmed the findings for the derivation group.

Hepatic gadoxetic acid uptake as a measure of diffuse liver disease: Where are we?

MRI has emerged as the most comprehensive noninvasive diagnostic tool for focal liver lesions and diffuse hepatobiliary disorders. The introduction of hepatobiliary contrast agents, most notably gadoxetic acid (GA), has expanded the role of MRI, particularly in the functional imaging of chronic liver diseases, such as nonalcoholic fatty liver disease (NAFLD). GA-enhanced MRI (GA-MRI) may help to distinguish between the two subgroups of NAFLD, simple steatosis and nonalcoholic steatohepatitis. Furthermore, GA-MRI can be used to stage fibrosis and cirrhosis, predict liver transplant graft survival, and preoperatively estimate the risk of liver failure should major resection be undertaken. The amount of GA uptake can be estimated, using static images, by the relative liver enhancement, hepatic uptake index, and relaxometry of T1-mapping during the hepatobiliary phase. On the contrary, the hepatic extraction fraction and liver perfusion can be measured on dynamic imaging. Importantly, there is currently no clear consensus as to which of these MR-derived parameters is the most suitable for assessing liver dysfunction. This review article aims to describe the current role of GA-enhanced MRI in quantifying liver function, primarily in diffuse hepatobiliary disorders. LEVEL OF EVIDENCE: 3 J. Magn. Reson. Imaging 2017;45:646-659.

The diagnostic efficacy of quantitative liver MR imaging with diffusion-weighted, SWI, and hepato-specific contrast-enhanced sequences in staging liver fibrosis--a multiparametric approach.

PURPOSE: To assess the diagnostic efficacy of multiparametric MRI using quantitative measurements of the apparent diffusion coefficient (ADC) of the liver parenchyma on diffusion-weighted imaging (DWI), signal intensity (SI) on susceptibility-weighted imaging (SWI), and gadoxetic acid-enhanced T1-weighted imaging during the hepatobiliary phase for the staging of liver fibrosis. MATERIALS AND METHODS: Seventy-seven patients underwent a 3T MRI examination, including DWI/SWI sequences and gadoxetic acid-enhanced T1-weighted MRI. Liver fibrosis according to liver biopsy was staged using the Metavir fibrosis score: F0 (n = 21, 27.3%); F1 (n = 7, 9.1%); F2 (n = 8, 10.4%); F3 (n = 12, 15.6%); and F4 (n = 29, 37.7%). SI of the liver was defined using region-of-interest measurements to calculate the ADC values, the relative enhancement (RE) in the hepatobiliary phase, and the liver-to-muscle ratio (LMR) measurements for SWI. RESULTS: The values of RE, LMR, and ADC measurements were statistically significantly different among the five fibrosis stages (p < 0.004). Combining the three parameters in a multiparametric approach, the AUC for detecting F1 stage or greater (≥ F1) was 94%, for F2 or greater (≥F2) was 95%, for F3 or greater (≥F3) was 90%, and for stage F4 was 93%. CONCLUSIONS: Multiparametric MRI is an efficient non-invasive diagnostic tool for the staging of liver fibrosis. KEY POINTS: • Multiparametric MRI has high accuracy in predicting moderate or greater liver fibrosis. • Relative enhancement post- gadoxetic acid is an independent predictor of liver fibrosis. • Liver SWI signal intensity and ADC values enhance the diagnostic ability.