Small hepatocellular carcinoma: ultrasonographic findings and histopathologic correlation

Development of novel deep multimodal representation learning-based model for the differentiation of liver tumors on B-mode ultrasound images

BACKGROUND AND AIM

Recently, multimodal representation learning for images and other information such as numbers or language has gained much attention. The aim of the current study was to analyze the diagnostic performance of deep multimodal representation model-based integration of tumor image, patient background, and blood biomarkers for the differentiation of liver tumors observed using B-mode ultrasonography (US).

METHOD

First, we applied supervised learning with a convolutional neural network (CNN) to 972 liver nodules in the training and development sets to develop a predictive model using segmented B-mode tumor images. Additionally, we also applied a deep multimodal representation model to integrate information about patient background or blood biomarkers to B-mode images. We then investigated the performance of the models in an independent test set of 108 liver nodules.

RESULTS

Using only the segmented B-mode images, the diagnostic accuracy and area under the curve (AUC) values were 68.52% and 0.721, respectively. As the information about patient background and blood biomarkers was integrated, the diagnostic performance increased in a stepwise manner. The diagnostic accuracy and AUC value of the multimodal DL model (which integrated B-mode tumor image, patient age, sex, aspartate aminotransferase, alanine aminotransferase, platelet count, and albumin data) reached 96.30% and 0.994, respectively.

CONCLUSION

Integration of patient background and blood biomarkers in addition to US image using multimodal representation learning outperformed the CNN model using US images. We expect that the deep multimodal representation model could be a feasible and acceptable tool for the definitive diagnosis of liver tumors using B-mode US.

Artificial intelligence in the diagnosis and management of hepatocellular carcinoma

Despite recent improvements in therapeutic interventions, hepatocellular carcinoma is still associated with a poor prognosis in patients with an advanced disease at diagnosis. Recently, significant progress has been made in image recognition through advances in the field of artificial intelligence (AI) (or machine learning), especially deep learning. AI is a multidisciplinary field that draws on the fields of computer science and mathematics for developing and implementing computer algorithms capable of maximizing the predictive accuracy from static or dynamic data sources using analytic or probabilistic models. Because of the multifactorial and complex nature of liver diseases, the machine learning approach to integrate multiple factors would appear to be an advantageous approach to improve the likelihood of making a precise diagnosis and predicting the response of treatment and prognosis of liver diseases. In this review, we attempted to summarize the potential use of AI in the diagnosis and management of liver diseases, especially hepatocellular carcinoma.

Novel protective role of curcumin and taurine combination against experimental hepatocarcinogenesis

Hepatocarcinogenesis is a prerequisite to hepatocellular carcinoma (HCC), which is one of the most common cancers among humans. Therefore, it is important to search for agents that protect against hepatocarcinogenesis. The present study aimed to investigate the protective effects of a combination of taurine and curcumin against experimental hepatocarcinogenesis induced by diethyl nitrosamine (DENA) in a rat model. A total of 100 rats were divided into eight groups. Eight weeks following DENA injection and treatment with curcumin and taurine, the rats were sacrificed to obtain blood and hepatic tissue samples for the evaluation of various markers and histopathological observations. Serum levels of interleukin-2 (IL-2), interferon-γ (IFN-γ), α-fetoprotein (AFP) and α-L-fucosidase (AFU) were determined. Rats injected with DENA for eight weeks showed a high percentage of malignant changes in hepatic tissues, as well as a significant increases in the serum levels of AFP and AFU and significant reductions in the serum levels of IL-2 and IFN-γ. Treatment with curcumin and taurine markedly reduced the extent of malignant changes in the rat liver tissues, with their liver tissues showing patterns similar to that of the normal control rats. In addition, this combination resulted in normal serum levels of IL-2, IFN-γ, AFP and AFU. The results of the present study suggested that a combination of curcumin and taurine may be a novel prophylactic agent against hepatocarcinogenesis in high-risk groups exposed to chemical hepatocarcinogens.

Characteristics suggestive of focal Fatty sparing from liver malignancy on ultrasound in liver screening

OBJECTIVE

The objective of this study was to determine whether there are ultrasound characteristics that can be used to differentiate focal fatty sparing (FFS) from hepatocellular carcinoma and other liver malignancies in liver screening.

METHODS

Data of patients with FFS found at ultrasound were reviewed, 136 patients with FFS were included, and 112 patients with hepatocellular carcinoma and 65 patients with liver metastatic tumor (173 tumors) were selected as control group. Ultrasound and color Doppler characteristics of FFS and liver malignancies were studied. Some characteristics drawn from this study were used as reference to validate the liver malignancies and FFS, and sensitivity, specificity, and positive and negative predictive values were calculated.

RESULTS

Hyperechogenicity, isoechogenicity, heterogeneous echotexture, posterior acoustic shadowing, halo, hypervasculature presented exclusively in liver malignancies, posterior acoustic enhancement presented predominantly in liver malignancies, irregularly shaped morphology presented exclusively in FFS, arterial velocity, and maximal size of liver malignancies were significantly higher than those of FFS (P < 0.001). Halo presented in 109 of 285 liver malignancies (38.2%), but was absent in FFS (P < 0.001). The overall sensitivity, specificity, and positive and negative predictive values were 94.38%, 90.44%, and 95.39% and 88.49%, respectively.

CONCLUSIONS

Combination of hyperechogenicity, isoechogenicity, heterogeneous echotexture, posterior acoustic shadowing, posterior acoustic enhancement, halo appearance, and hypervascularity with high-speed artery has high sensitivity, specificity, and positive and negative predictive values for distinguishing liver malignancy and FFS.

The 3' UTR variants in the GRP78 are not associated with overall survival in resectable hepatocellular carcinoma

Background

Elevated glucose-regulated protein 78 (GRP78) levels in tissues have been known to be related with poor prognosis in hepatocellular carcinoma (HCC) patients. Though the variants in the 3′ untranslated region (UTR) of GRP78 gene were not associated with HCC risk, we wonder whether these polymorphisms affect survival of HCC patients.

Methodology/Principal Findings

Blood samples of HCC patients were maintained in our specimen bank between 1996 to 2003. DNA from 576 unrelated and resectable patients with HCC was typed for rs16927997 (T>C), rs1140763 (T>C) and rs12009 (T>C) by TaqMan assays. The Kaplan-Meier method and log-rank test were used to estimate overall survival. Linkage disequilibrium (LD) analysis identified a total of 3 haplotypes and 6 diplotypes in this region. The distribution of haplotype was not related to the clinical characteristics. Univariate analysis showed that the allele, genotype, haplotype and diplotype did not effect the survival. None of the clinical features show a significant association (P_correced>0.05) with overall patient outcome in multiple comparisons.

Conclusions/Significance

There is no noteworthy influence of 3′ UTR variants in the GRP78 on prognosis of resectable HCC in the Chinese population.

Clinical evaluation of serum alpha-fetoprotein-IgM immune complexes on the diagnosis of primary hepatocellular carcinoma

We evaluated clinical significance of serum alpha-fetoprotein (AFP)-IgM immune complexes, in comparison with free AFP, on the diagnosis of primary hepatocellular carcinoma (HCC). Serum levels of AFP-IgM immune complexes and free AFP were determined by the ELISA method and electrochemiluminescence, respectively, in 103 healthy controls, 74 patients suffering from primary HCC, 27 patients suffering from liver cirrhosis, and 63 patients suffering from chronic hepatitis. The best cut-off value of AFP-IgM and free AFP for diagnosis of primary HCC were 300 AU/mL and 10 microg/L respectively, according to the area under the curve (AUC) in this study. The sensitivity of AFP-IgM and free AFP were 64.9 and 79.7%, and the specificity were 75.6 and 80.3%, respectively, when all cases of primary HCC were analyzed, and the AUC of free AFP was larger than that of AFP-IgM (0.85 vs. 0.72, Z=3.21). However, in case of primary HCC at early stages (stages I and II) were analyzed, the AUC of AFP-IgM was larger than that of free AFP (0.91 vs. 0.82, Z=1.73), which demonstrated that the sensitivity of AFP-IgM and free AFP were 94.4 and 72.2%, and the specificity were 81.9 and 79.9%, respectively. When both AFP-IgM and free AFP were positive, the specificity of diagnosis of primary HCC was 89.1%, and the efficacy was 79.0%. It is concluded that either sensitivity or specificity of serum level of AFP-IgM immune complexes was higher than that of free AFP in the diagnosis of primary HCC at early stages. As there was false positive AFP-IgM existed in the patients suffering from cirrhosis and chronic hepatitis, the combination of free AFP and AFP-IgM could significantly increase specificity and decrease false negative and/or false positive in the primary HCC at early stages.

[Imaging nodules within cirrhotic liver: how do I do it?]

Cirrhosis is a chronic liver disease characterized by the presence of diffuse parenchymal necrosis, reactive fibrosis and nodular regeneration. These regenerative nodules may evolve into dysplastic nodules and finally nodules of hepatocellular carcinoma (HCC). Improved survival of cirrhotic patients with HCC depends on eligibility to liver transplantation. The purpose of this paper is to review the imaging features of liver nodules within cirrhotic liver and to propose the imaging strategies when considering the possibility of liver transplantation.