Feasibility of semiquantitative liver perfusion assessment by ferucarbotran bolus injection in double-contrast hepatic MRI

Application of A U-Net for Map-like Segmentation and Classification of Discontinuous Fibrosis Distribution in Gd-EOB-DTPA-Enhanced Liver MRI

We aimed to evaluate whether U-shaped convolutional neuronal networks can be used to segment liver parenchyma and indicate the degree of liver fibrosis/cirrhosis at the voxel level using contrast-enhanced magnetic resonance imaging. This retrospective study included 112 examinations with histologically determined liver fibrosis/cirrhosis grade (Ishak score) as the ground truth. The T1-weighted volume-interpolated breath-hold examination sequences of native, arterial, late arterial, portal venous, and hepatobiliary phases were semi-automatically segmented and co-registered. The segmentations were assigned the corresponding Ishak score. In a nested cross-validation procedure, five models of a convolutional neural network with U-Net architecture (nnU-Net) were trained, with the dataset being divided into stratified training/validation (n = 89/90) and holdout test datasets (n = 23/22). The trained models precisely segmented the test data (mean dice similarity coefficient = 0.938) and assigned separate fibrosis scores to each voxel, allowing localization-dependent determination of the degree of fibrosis. The per voxel results were evaluated by the histologically determined fibrosis score. The micro-average area under the receiver operating characteristic curve of this seven-class classification problem (Ishak score 0 to 6) was 0.752 for the test data. The top-three-accuracy-score was 0.750. We conclude that determining fibrosis grade or cirrhosis based on multiphase Gd-EOB-DTPA-enhanced liver MRI seems feasible using a 2D U-Net. Prospective studies with localized biopsies are needed to evaluate the reliability of this model in a clinical setting.

Liver function correlates with liver-to-portal vein contrast ratio during the hepatobiliary phase with Gd-EOB-DTPA-enhanced MR at 3 Tesla

PURPOSES

To determine if liver-to-portal vein contrast ratio (LPC) correlates with liver function in patients with hepatitis B virus (HBV)-related cirrhosis on gadolinium-ethoxybenzyl-diethylenetriamine penta-acetic acid (Gd-EOB-DTPA)-enhanced MR imaging.

METHODS

A total of 92 patients with normal (n = 20) or HBV-related cirrhotic livers graded by Child-Pugh class A (n = 50), B (n = 17) or C (n = 5) who underwent Gd-EOB-DTPA-enhanced 3Tesla MR imaging were retrospectively reviewed. LPC was defined as the signal intensity ratio of liver parenchyma to portal vein on hepatobiliary phase (HBP) acquired at 20 min, and it was compared between normal and cirrhotic livers. The correlation between LPC and hepatic function parameters at HBP after injection was quantitatively analyzed as well.

RESULTS

LPC differed between normal and cirrhotic livers significantly (P < 0.001). LPC constantly and significantly decreased from normal to cirrhotic livers with Child-Pugh class C at HBP imaging (P < 0.001). Multiple regression analysis revealed that total bilirubin (P = 0.011), albumin (P < 0.001), and platelet count (P = 0.007) were independent predictors of LPC at HBP imaging. A receiver operating characteristic (ROC) curve analysis revealed that the optimal cutoff value for LPC to distinguish normal group from cirrhotic groups was 2.05 (AUC 0.98) with a sensitivity of 84.1% and a specificity of 100%.

CONCLUSION

The level of LPC on Gd-EOB-DTPA MR imaging can efficaciously indicate the severity of liver function in patients with HBV-related cirrhosis and was correlated with liver function parameters significantly. It might be used as an alternative imaging biomarker for assessing liver function.

Liver fibrosis and Gd-EOB-DTPA-enhanced MRI: A histopathologic correlation

Gadolinium ethoxybenzyl-diethylenetriaminepentaacetic acid (Gd-EOB-DTPA) is a hepatocyte-specific MRI contrast agent. Because the hepatic uptake of Gd-EOB-DTPA depends on the integrity of the hepatocyte mass, this uptake can be quantified to assess liver function. We report the relationship between the extent of Gd-EOB-DTPA uptake and the degree of liver fibrosis. T1-weighted volume-interpolated breath-hold examination (VIBE) sequences with fat suppression were acquired before and 20 minutes after contrast injection. Strong correlations of the uptake characteristics of Gd-EOB-DTPA with the relative enhancement (RE) of the liver parenchyma and the grade of fibrosis/cirrhosis, classified using the Ishak scoring system, were observed. The subdivisions between the grades of liver fibrosis based on RE were highly significant for all combinations, and a ROC revealed sensitivities ≥82% and specificities ≥87% for all combinations. MR imaging is a satisfactorily sensitive method for the assessment of liver fibrosis/cirrhosis.

A Multilevel Modeling Framework to Study Hepatic Perfusion Characteristics in Case of Liver Cirrhosis

Liver cirrhosis represents the end-stage of different liver disorders, progressively affecting hepatic architecture, hemodynamics, and function. Morphologically, cirrhosis is characterized by diffuse fibrosis, the conversion of normal liver architecture into structurally abnormal regenerative nodules and the formation of an abundant vascular network. To date, the vascular remodeling and altered hemodynamics due to cirrhosis are still poorly understood, even though they seem to play a pivotal role in cirrhogenesis. This study aims to determine the perfusion characteristics of the cirrhotic circulation using a multilevel modeling approach including computational fluid dynamics (CFD) simulations. Vascular corrosion casting and multilevel micro-CT imaging of a single human cirrhotic liver generated detailed datasets of the hepatic circulation, including typical pathological characteristics of cirrhosis such as shunt vessels and dilated sinusoids. Image processing resulted in anatomically correct 3D reconstructions of the microvasculature up to a diameter of about 500 μm. Subsequently, two cubic samples (150 × 150 × 150 μm3) were virtually dissected from vascularized zones in between regenerative nodules and applied for CFD simulations to study the altered cirrhotic microperfusion and permeability. Additionally, a conceptual 3D model of the cirrhotic macrocirculation was developed to reveal the hemodynamic impact of regenerative nodules. Our results illustrate that the cirrhotic microcirculation is characterized by an anisotropic permeability showing the highest value in the direction parallel to the central vein (kd,zz = 1.68 × 10−13 m2 and kd,zz = 7.79 × 10−13 m2 for sample 1 and 2, respectively) and lower values in the circumferential (kd,ϑϑ = 5.78 × 10−14 m2 and kd,ϑϑ = 5.65 × 10−13 m2 for sample 1 and 2, respectively) and radial (kd,rr = 9.87 × 10−14 m2 and kd,rr = 5.13 × 10−13 m2 for sample 1 and 2, respectively) direction. Overall, the observed permeabilities are markedly higher compared to a normal liver, implying a locally decreased intrahepatic vascular resistance (IVR) probably due to local compensation mechanisms (dilated sinusoids and shunt vessels). These counteract the IVR increase caused by the presence of regenerative nodules and dynamic contraction mechanisms (e.g., stellate cells, NO-concentration, etc.). Our conceptual 3D model of the cirrhotic macrocirculation indicates that regenerative nodules severely increase the IVR beyond about 65 vol. % of regenerative nodules. Numerical modeling allows quantifying perfusion characteristics of the cirrhotic macro- and microcirculation, i.e., the effect of regenerative nodules and compensation mechanisms such as dilated sinusoids and shunt vessels. Future research will focus on the development of models to study time-dependent degenerative adaptation of the cirrhotic macro- and microcirculation.

A fast nonlinear regression method for estimating permeability in CT perfusion imaging

Blood-brain barrier damage, which can be quantified by measuring vascular permeability, is a potential predictor for hemorrhagic transformation in acute ischemic stroke. Permeability is commonly estimated by applying Patlak analysis to computed tomography (CT) perfusion data, but this method lacks precision. Applying more elaborate kinetic models by means of nonlinear regression (NLR) may improve precision, but is more time consuming and therefore less appropriate in an acute stroke setting. We propose a simplified NLR method that may be faster and still precise enough for clinical use. The aim of this study is to evaluate the reliability of in total 12 variations of Patlak analysis and NLR methods, including the simplified NLR method. Confidence intervals for the permeability estimates were evaluated using simulated CT attenuation-time curves with realistic noise, and clinical data from 20 patients. Although fixating the blood volume improved Patlak analysis, the NLR methods yielded significantly more reliable estimates, but took up to 12 × longer to calculate. The simplified NLR method was ∼4 × faster than other NLR methods, while maintaining the same confidence intervals (CIs). In conclusion, the simplified NLR method is a new, reliable way to estimate permeability in stroke, fast enough for clinical application in an acute stroke setting.

Impact of liver cirrhosis on liver enhancement at Gd-EOB-DTPA enhanced MRI at 3 Tesla

PURPOSE

The purpose of this study was to assess differences in enhancement effects of liver parenchyma between normal and cirrhotic livers on dynamic, Gd-EOB-DTPA enhanced MRI at 3T.

MATERIALS AND METHODS

93 patients with normal (n=54) and cirrhotic liver (n=39; Child-Pugh class A, n=18; B, n=16; C, n=5) underwent contrast-enhanced MRI with liver specific contrast media at 3T. T1-weighted volume interpolated breath hold examination (VIBE) sequences with fat suppression were acquired before contrast injection, in the arterial phase (AP), in the late arterial phase (LAP), in the portal venous phase (PVP), and in the hepatobiliary phase (HBP) after 20 min. The relative enhancement (RE) of the signal intensity of the liver parenchyma was calculated for all phases.

RESULTS

Mean RE was significantly different among all evaluated groups in the hepatobiliary phase and with increasing severity of liver cirrhosis, a decreasing, but still significant reduction of RE could be shown. Phase depending changes of RE for each group were observed. In case of non-cirrhotic liver or Child-Pugh Score A cirrhosis mean RE showed a significant increase between AP, LAP, PVP and HBP. For Child-Pugh B+C cirrhosis RE increased until PVP, however, there was no change in case of B cirrhosis (p=0.501) and significantly reduced in case of C cirrhosis (p=0.043) during HBP.

CONCLUSION

RE of liver parenchyma is negatively affected by increased severity of liver cirrhosis, therefore diagnostic value of HBP could be limited in case of Child Pugh B+C cirrhosis.