Deep learning for assessing liver fibrosis based on acoustic nonlinearity maps: an in vivo study of rabbits.

This study aimed to assess liver fibrosis in rabbits by deep learning models based on acoustic nonlinearity maps. Injection of carbon tetrachloride was used to induce liver fibrosis. Acoustic nonlinearity maps, which were built by data of echo signals, were used as input data for deep learning model. Convolutional neural network (CNN), CNN combined with support vector machine (SVM), CNN combined with random forest and CNN combined with logistic regression were used as deep learning model. Nested 10-fold cross-validation was used to search hyperparameters and evaluate performance of models. Histologic examination of liver specimens of the rabbits was performed to evaluate the fibrosis stage. Receiver operator characteristic curve and area under curve (AUC) were used for estimating the probability of the correct prediction of liver fibrosis stages. A total of 600 acoustic nonlinearity maps were used. Model of CNN combined with SVM demonstrated the best diagnostic performance compared with all other methods for diagnosis of significant fibrosis (≥F2, AUC = 0.82), advanced fibrosis (≥F3, AUC = 0.88) and cirrhosis (F4, AUC = 0.90). Model of CNN showed the second highest AUCs. The deep learning model based on acoustic nonlinearity maps demonstrated potential for evaluation of liver fibrosis.

Can Location of Stiffness Measurement Impact Spleen 2-Dimensional Shear Wave Elastography Measurement?

ABSTRACT: Ultrasound-based spleen elastography is a promising surrogate to predict portal hypertension noninvasively. In contrast to defined standards for liver stiffness measurement, the standardized examination procedures for 2-dimensional (2D) shear wave elastography spleen elastography have not been established yet. The aim was to investigate the impact of location of stiffness measurement on 2D shear wave elastography spleen stiffness measurement (SSM). Patients with splenomegaly were enrolled. Both B-mode ultrasound and elastography of spleen were performed. For SSM, 3 regions were chosen for spleen measurement: lower pole region, central region, and the region between lower pole and center. Mean SSM value, success rate, and reliability predicators (standard deviation, standard deviation/mean, size of region of interest) were assessed. A total of 124 patients were included. For mean SSM value, there were no significant differences among 3 regions. Spleen stiffness measurement success rate in lower pole region, central region, and the region between them was 63.7% (79), 91.1% (113), and 78.2% (97), respectively. The success rate in the central region was significantly higher than that in the other 2 regions (P < 0.05). Reliability in the central region was also highest among the 3 regions. Location of stiffness measurement has a limited effect on SSM. Changing location of measurement will not influence mean stiffness value in spleen.

Reliability of Transient Elastography-Based Liver Stiffness for Diagnosing Portal Hypertension in Patients with Alcoholic Liver Disease: A Diagnostic Meta-Analysis with Specific Cut-Off Values.

BACKGROUND: Transient elastography-based liver stiffness value (TE-LSV) has been studied for the diagnosis of portal hypertension. Liver stiffness is influenced by the disease etiology. We aimed to perform a meta-analysis to determine the performance of TE-LSV for diagnosing portal hypertension in patients with alcoholic liver disease (ALD). METHODS: We searched PubMed, Web of Science, Ovid and Cochrane library. A bivariate model was used to compute sensitivity and specificity. A random effects model was used to pool diagnostic odds ratios. RESULTS: 9 studies with 679 patients were included. The pooled sensitivity and specificity based on a cut-off value around 21.8 kPa for clinically significant portal hypertension (CSPH) were 0.89 (95 % confidence interval (CI), 0.83-0.93) and 0.71(95 % CI, 0.64-0.78), respectively. For severe portal hypertension (SPH), the pooled sensitivity and specificity for a cut-off value around 29.1 kPa were 0.88 (95 % CI, 0.83-0.92) and 0.74 (95 % CI, 0.67-0.81), respectively. CONCLUSION: TE-LSV showed good performance for diagnosing portal hypertension in patients with ALD. The optimal cut-off value for CSPH and SPH was around 21.8 kPa and 29.1 kPa, respectively, and these two cut-off values showed good sensitivity and modest specificity. The etiology should be clear before using TE-LSV for portal hypertension.

Comparison of three cut-offs to diagnose clinically significant portal hypertension by liver stiffness in chronic viral liver diseases: a meta-analysis.

BACKGROUND: Transient elastography-based liver stiffness value (TE-LSV) has been investigated for assessing clinically significant portal hypertension (CSPH). The aetiology of CSPH is an important factor determining TE-LSV. There is insufficient evidence for selecting cut-off values. AIMS: This study performed a meta-analysis to compare the three most widely used cut-off values (around 13.6 kPa, 18 kPa and 22kPa) of TE-LSV for the diagnosis of CSPH in patients with chronic viral liver disease. METHODS: The PubMed, Ovid, Web of Science and Cochrane Library databases were searched. Diagnostic data for cut-off values around 13.6 kPa, 18 kPa and 22 kPa in each included study were extracted. The bivariate model was performed to estimate pooled sensitivity, specificity, positive likelihood ratio (LR+) and negative likelihood ratio (LR-). RESULTS: Eleven studies assessing 910 patients were included in this meta-analysis. Pooled sensitivities of cut-off values around 13.6 kPa, 18 kPa and 22 kPa were 0.96 (95% CI 0.93-0.97), 0.85 (0.81-0.89) and 0.74 (0.66-0.80), respectively; pooled specificities were 0.60 (0.47-0.75), 0.80 (0.71-0.87) and 0.94 (0.86-0.97), respectively. Pooled LR+ values were 2.4 (1.6-3.7), 4.4 (2.9-6.8) and 11.5 (5.5-23.5) for cut-off values around 13.6 kPa, 18 kPa and 22 kPa, respectively, for pooled LR- values of 0.07 (0.04-0.13), 0.17 (0.12-0.25) and 0.28 (0.22-0.36), respectively. CONCLUSION: Cut-off values around 13.6 kPa (high sensitivity) and 22 kPa (high specificity) could be used as screening and confirmation tools, respectively, in the diagnosis of CSPH. Overall, the cut-off value around 22 kPa showed the best performance. KEY POINTS: Transient elastography-based liver stiffness could be used to diagnose portal hypertension. Comparison of certain cut-off values would provide more information for clinical decision-making. Cut-off around 13.6 kPa was able to exclude clinically significant portal hypertension (CSPH) effectively. Cut-off around 22 kPa was able to confirm CSPH effectively.

Could Ultrasound Elastography Reflect Liver Function?

The purpose of this study was to investigate whether ultrasound elastography reflects liver function reserve relative to liver fibrosis histology. Sixty-five New Zealand rabbits were divided into an experimental group (n = 45) and a control group (n = 20). In the experimental group, liver fibrosis (F1-F4) was induced by subcutaneous injection of carbon tetrachloride. Point shear wave elastography and the indocyanine green (ICG) elimination test were performed for the two groups at 4-wk intervals for 56 wk. The liver stiffness value (LSV) and the ICG retention rate at 15 min (ICGR15) were obtained, and the correlation between them was investigated. The median LSVs of stages F0-F4 were 3.92 kPa (1.91-8.53 kPa), 5.02 kPa (2.39-8.91 kPa), 7.87 kPa (5.21-12.26 kPa), 12.83 kPa (5.92-16.79 kPa) and 16.64 kPa (9.76-29.50 kPa), respectively. The median ICGR15 values of stages F0-F4 were 8.7% (4.8%-15.6%), 10.8% (5.6%-20.3%), 19.2% (12.3%-26.7%), 31.0% (20.9%-41.0%) and 45.6% (22.1%-60.9%). There were significant differences in LSVs and ICGR15 values among the different stages of liver fibrosis (p <0.01). A positive correlation was observed between LSV and ICGR15 (r = 0.7497, p < 0.0001). A strong correlation was observed between liver stiffness and liver function reserve, indicating ultrasound elastography may reflect liver function reserve in different degrees of liver fibrosis.

Performance of spleen stiffness measurement in prediction of clinical significant portal hypertension: A meta-analysis.

AIM: Our purpose was to evaluate the correlation between spleen stiffness (SS) measured by ultrasound-based elastography and hepatic venous pressure gradient (HVPG) and assess the accuracy of SS in detecting clinical significant portal hypertension (CSPH) and severe portal hypertension. METHOD: Nine studies were included from thorough literature research and selection processes. A random model was used to analyze the correlation between HVPG and SS. We adopted the bivariate mixed effects model to assess the diagnostic performance. RESULTS: Regarding to correlation between SS and HVPG, the summary correlation coefficient was 0.72 (95% confidence interval [CI], 0.63-0.80). In detection of CSPH, the sensitivity, specificity, AUC and DOR were: 0.88 (0.70-0.96), 0.84 (0.72-0.92), 0.92 (0.89-0.94) and 38 (17-84) for CSPH, respectively; and 0.92 (0.82-0.96), 0.79 (0.72-0.85), 0.87 (0.84-0.90) and 41 (17-100) for severe portal hypertension, respectively. CONCLUSION: Correlation between SS and HVPG was good. Although SS showed good sensitivity and specificity, the different cut-off values and techniques among studies might limit the impact of our results on clinical practice. Therefore, more high-quality prospective studies are required to evaluate the role of SS in predicting portal hypertension.

Assessment of liver fibrosis by ultrasound elastography and contrast-enhanced ultrasound: a randomized prospective animal study.

This study aimed to assess liver fibrosis by contrast-enhanced ultrasound (CEUS) and point shear-wave elastography (pSWE) in rabbits and compare the performance of the two techniques. Eighty rabbits were divided into experimental (n=60) and control group (n=20). In the experimental group, liver fibrosis (F1-F4) was induced by subcutaneous injection of carbon tetrachloride. CEUS and pSWE of the liver was performed for the two groups at a 4-week interval for 40 weeks. The portal vein rise time (PV-RT), time to peak (PV-TTP), mean transit time (PV-MTT) and the maximum signal intensity (PV-Imax) were analyzed with time-intensity curves (TICs). Liver stiffness value (LSV) was obtained through pSWE. Histologic examination of liver specimens of the rabbits was performed to evaluate the fibrosis stage. PV-RT, PV-TTP, PV-Imax and LSV were significantly different among five liver fibrosis stages (F0-F4) (P<0.01). PV-Imax and LSV displayed better diagnostic performance than PV-RT, PV-TTP, PV-MTT. For diagnosing≥F1 stage fibrosis, the area under the receiver operating characteristic curve (AUROC) of PV-Imax was 0.870, which was similar to that of LSV 0.874 (P=0.94). For diagnosing ≥F2, ≥F3 and ≥F4 stage fibrosis, the AUROC of PV-Imax and LSV was 0.845 vs. 0.956 (P=0.04), 0.789 vs. 0.954 (P=0.01) and 0.707 vs. 0.933 (P=0.03). Both CEUS and pSWE had the potential to be complementary imaging tools in the evaluation of liver fibrosis. The performance of pSWE may be better than CEUS.