MRI evaluation of fatty liver in day to day practice: Quantitative and qualitative methods

Post-COVID-19 Liver Injury: Comprehensive Imaging With Multiparametric Ultrasound

OBJECTIVES

This study aimed to define patterns of liver injury after severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection using multiparametric ultrasound (mpUS) in a variable patient population with differing severities of COVID-19.

METHODS

Ninety patients were enrolled into the study: 56 had SARS-CoV-2 3-9 months prior to enrolment; 34 served as a clinically healthy control group. All patients underwent an mpUS evaluation of the liver (elastography, dispersion and attenuation imaging). Seventy-six patients had abdominal magnetic resonance (MR) and noncontrast enhanced thoracic computed tomography (CT) scans performed at the same day. All patients were screened for biochemical markers of liver injury.

RESULTS

Liver elasticity, viscosity, and steatosis values were significantly altered in patients after COVID-19, with particularly higher fibrosis scores compared to the control group (P < .001). Increased biochemical markers of liver injury correlated with changes in mpUS (P < .05), but not with findings on CT or MR findings. Seventeen of 34 hospitalized patients had a moderate or severe course of the disease course with more pronounced changes in mpUS. Increased body mass index was found to influence liver injury and correlated with more severe forms of COVID-19 (P < .001).

CONCLUSIONS

COVID-19 can cause liver injury observable using mpUS. More severe forms of COVID-19 and patient obesity are related to increased values of liver damage observed. In comparison to MRI and CT, mpUS appears to be more sensitive to involvement of liver parenchyma. Further research is warranted to establish this promising method for evaluating post-COVID-19 liver involvement in the aftermath of the pandemic.

Correlation between incidental fat deposition in the liver and pancreas in asymptomatic individuals

PURPOSE

To explore the utility of two different fat quantification methods in the liver and pancreas and to test the accuracy of multi-echo Dixon as a single sequence in detecting early stage of fat deposition.

METHODS

58 healthy potential liver donors underwent abdominal 3T MRI, prospectively. Single-voxel MR Spectroscopy (MRS), dual-echo Dixon, and multi-echo Dixon were performed. Two independent readers obtained proton density fat fraction (PDFF) of the liver and pancreas by placing ROIs on the 2 Dixon sequences. Correlation between the two PDFF measurements was assessed in the liver and pancreas. Values in the liver were also compared to those obtained by MRS.

RESULTS

PDFF in the liver was 6.3 ± 4.2%, 5.5 ± 3.9%, and 5.1 ± 4.1% by MRS, dual-echo Dixon, and multi-echo Dixon, respectively. Dual-echo Dixon and multi-echo Dixon showed good correlation in PDFF quantification of the liver (r = 0.82, p < 0.0005). Multi-echo Dixon showed a good correlation (r = 0.72, p = 0.0005) between the fat measured in the liver and in the pancreas. To differentiate between normal (PDFF ≤ 6%) and mild fat deposition (PDFF: 6-33%) in the liver, analysis showed sensitivity, specificity, and accuracy of 74%, 81%, and 80% for dual-echo Dixon and 85%, 96%, and 89% for multi-echo Dixon, respectively. Mean PDFF in the pancreas was 7.2 ± 2.8% and 6.7 ± 3.3%, by dual-echo and multi-echo Dixon, respectively. Dual-echo Dixon and multi-echo Dixon showed good correlation in PDFF quantification of the pancreas (r = 0.58, p < 0.0005).

CONCLUSION

Multi-echo Dixon in liver has high accuracy in distinguishing between subjects with normal liver fat and those with mildly elevated liver fat. Multi-echo Dixon can be used to screen for early fat deposition in the liver and pancreas.

Symtosis: A liver ultrasound tissue characterization and risk stratification in optimized deep learning paradigm

Background and Objective Fatty Liver Disease (FLD) - a disease caused by deposition of fat in liver cells, is predecessor to terminal diseases such as liver cancer. The machine learning (ML) techniques applied for FLD detection and risk stratification using ultrasound (US) have limitations in computing tissue characterization features, thereby limiting the accuracy. Methods Under the class of Symtosis for FLD detection and risk stratification, this study presents a Deep Learning (DL)-based paradigm that computes nearly seven million weights per image when passed through a 22 layered neural network during the cross-validation (training and testing) paradigm. The DL architecture consists of cascaded layers of operations such as: convolution, pooling, rectified linear unit, dropout and a special block called inception model that provides speed and efficiency. All data analysis is performed in optimized tissue region, obtained by removing background information. We benchmark the DL system against the conventional ML protocols: support vector machine (SVM) and extreme learning machine (ELM). Results The liver US data consists of 63 patients (27 normal/36 abnormal). Using the K10 cross-validation protocol (90% training and 10% testing), the detection and risk stratification accuracies are: 82%, 92% and 100% for SVM, ELM and DL systems, respectively. The corresponding area under the curve is: 0.79, 0.92 and 1.0, respectively. We further validate our DL system using two class biometric facial data that yields an accuracy of 99%. Conclusion DL system shows a superior performance for liver detection and risk stratification compared to conventional machine learning systems: SVM and ELM.

The chloroform extract of Cyclocarya paliurus attenuates high-fat diet induced non-alcoholic hepatic steatosis in Sprague Dawley rats

BACKGROUND

Hepatic steatosis (HS) is the early stage of nonalcoholic fatty liver disease which is caused by impaired hepatic lipid homeostasis. Cyclocarya paliurus, an herbal tea consumed in China, has been demonstrated to ameliorate abnormal lipid metabolism for the treatment of metabolic diseases.

PURPOSE

We aimed to investigate the regulative effect of chloroform extract from Cyclocarya paliurus (ChE) on treatment of HS, as well as key factors involved in hepatic lipid metabolism.

STUDY DESIGN

Sprague Dawley rats were fed with high-fat diet (HFD) for 6 weeks to induce HS and treated with or without ChE by gavage for 4 weeks.

METHODS

The body weight, relative liver weight and liver fat content were measured. Serum and liver total cholesterol, triglyceride and non-esterified fatty acids, as well as hepatic malonaldehyde levels were accessed by biochemical methods. Serum and liver TNF-α levels were quantified by ELISA kit. Histologic analysis and ¹H-MRS study were performed to evaluate HS level. RT-PCR and Western blot were also applied to observe the expression changes of key factors involved in hepatic lipid intake, synthesis, utilization and export.

RESULTS

ChE significantly decreased the rats' body weight, serum lipid and TNF-α level. ChE also reduced their relative liver weight, liver fat content, hepatic oxidative products and TNF-α level. Hepatic steatosis in HFD-fed rats was effectively regressed after 2-weeks administration of ChE. Moreover, ChE treatment remarkably reduced HFD-induced high expression level of fatty acid synthesis genes (including sterol-regulatory element-binding protein 1, acetyl-CoA carboxylase 1 and fatty acid synthase). However, it had no effect on mRNA expression of some genes involved in lipid uptake, β-oxidation and lipid outflow.

CONCLUSION

ChE exerted a promising regression effect on HS due to a reduced level of serum non-esterified fatty acids which might lead to a decrease in the amount of lipid taken in by the liver, as well as owing to the inhibition of hepatic lipid de novo synthesis to reduce liver lipid production.