Clinical Application of Quantitative MR Imaging in Nonalcoholic Fatty Liver Disease

Severe hepatic steatosis promotes increased liver stiffness in the early stages of metabolic dysfunction-associated steatotic liver disease

BACKGROUND & AIMS

The predictors of progression from steatosis to more advanced stages of metabolic dysfunction-associated steatotic liver disease (MASLD) remain unclear. We evaluated the association between the quantity of hepatic steatosis and longitudinal changes in liver stiffness measurements (LSMs) using magnetic resonance elastography (MRE) in patients with MASLD.

METHODS

We retrospectively analysed patients with MASLD who underwent at least two serial MRE and magnetic resonance imaging-based proton density fat fraction (MRI-PDFF) examinations at least 1 year apart. Fine-Gray competitive proportional hazard regression was used to identify LSM progression and regression factors.

RESULTS

A total of 471 patients were enrolled. Factors linked to LSM progression were steatosis grade 3 (MRI-PDFF ≥17.1%, adjusted hazard ratio [aHR] 2.597; 95% confidence interval [CI] 1.483-4.547) and albumin-bilirubin grade 2 or 3 (aHR 2.790; 95% CI 1.284-6.091), while the only factor linked to LSM regression was % decrease rate of MRI-PDFF ≥5% (aHR 2.781; 95% CI 1.584-4.883). Steatosis grade 3 correlated with a higher incidence rate of LSM progression than steatosis grade 1 (MRI-PDFF <11.3%) in patients with LSM stage 0 (<2.5 kilopascal [kPa]), and a % annual decrease rate of MRI-PDFF ≥5% correlated with a higher incidence rate of LSM regression than that of MRI-PDFF >-5% and <5% in patients with LSM stage 1 or 2-4 (≥2.5 kPa).

CONCLUSIONS

Severe hepatic steatosis was linked to significant LSM progression in patients with MASLD and low LSM (<2.5 kPa).

Advanced fibrosis leads to overestimation of steatosis with quantitative ultrasound in individuals without hepatic steatosis

PURPOSE

The effect of hepatic fibrosis stage on quantitative ultrasound based on the attenuation coefficient (AC) for liver lipid quantification is controversial. The objective of this study was to determine how the degree of fibrosis assessed by magnetic resonance (MR) elastography affects AC based on the ultrasound-guided attenuation parameter according to the grade of hepatic steatosis, using magnetic resonance imaging (MRI)-derived proton density fat fraction (MRIderived PDFF) as the reference standard.

METHODS

Between February 2020 and April 2021, 982 patients with chronic liver disease who underwent AC and MRI-derived PDFF measurement as well as MR elastography were enrolled. Multiple regression was used to investigate whether AC was affected by the degree of liver stiffness.

RESULTS

AC increased as liver stiffness progressed in 344 patients without hepatic steatosis (P=0.009). In multivariable analysis, AC was positively correlated with skin-capsule distance (P<0.001), MR elastography value (P=0.037), and MRI-derived PDFF (P<0.001) in patients without hepatic steatosis. In 52 of 982 patients (5%), the correlation between AC and MRIderived PDFF fell outside the 95% confidence interval for the regression line slope. Patients with MRI-derived PDFF lower than their AC (n=36) had higher fibrosis-4 scores, albumin-bilirubin scores, and MR elastography values than patients with MRI-derived PDFF greater than their AC (n=16; P=0.018, P=0.001, and P=0.011, respectively).

CONCLUSION

AC is affected by liver fibrosis (MR elastography value ≥6.7 kPa) only in patients without hepatic steatosis (MRI-derived PDFF <5.2%). These values should be interpreted with caution in patients with advanced liver fibrosis.

Clinical applications of artificial intelligence in liver imaging

This review outlines the current status and challenges of the clinical applications of artificial intelligence in liver imaging using computed tomography or magnetic resonance imaging based on a topic analysis of PubMed search results using latent Dirichlet allocation. LDA revealed that "segmentation," "hepatocellular carcinoma and radiomics," "metastasis," "fibrosis," and "reconstruction" were current main topic keywords. Automatic liver segmentation technology using deep learning is beginning to assume new clinical significance as part of whole-body composition analysis. It has also been applied to the screening of large populations and the acquisition of training data for machine learning models and has resulted in the development of imaging biomarkers that have a significant impact on important clinical issues, such as the estimation of liver fibrosis, recurrence, and prognosis of malignant tumors. Deep learning reconstruction is expanding as a new technological clinical application of artificial intelligence and has shown results in reducing contrast and radiation doses. However, there is much missing evidence, such as external validation of machine learning models and the evaluation of the diagnostic performance of specific diseases using deep learning reconstruction, suggesting that the clinical application of these technologies is still in development.