A meta-analysis of the added value of diffusion weighted imaging in combination with contrast-enhanced magnetic resonance imaging for the diagnosis of small hepatocellular carcinoma lesser or equal to 2 cm

Diffusion weighted imaging (DWI) has been found to increase the sensitivity in the diagnosis of small hepatocellular carcinoma (HCC), although additional studies are required to confirm its value. The aim of the present study was to explore the diagnostic performance of DWI combined with contrast-enhanced magnetic resonance imaging (MRI) for small HCC by performing a meta-analysis. Literature databases (PubMed, Embase, Web of Science and Cochrane Library databases) were searched to identify studies reporting the sensitivity and specificity of MRI with DWI for the diagnosis of small HCCs. Pooled sensitivity and specificity were generated using a bivariate random effect model. Multilevel mixed-effects logistic regression analysis was used to examine the value of DWI combined with conventional MRI. A total of 837 small HCCs and 545 benign liver lesions from 10 studies were included. The overall sensitivity and specificity of DWI combined with contrast-enhanced MRI was 0.88 (95% CI, 0.80-0.93) and 0.90 (95% CI, 0.81-0.95), respectively. Compared with that in contrast-enhanced MRI, DWI with contrast-enhanced MRI had a significantly higher sensitivity for the diagnosis of small HCC (P=0.01) while there was no significant difference in the specificity (P=0.603). The present meta-analysis suggests that DWI combined with contrast-enhanced MRI may increase the sensitivity, whilst maintaining high specificity for the diagnosis of small HCCs with a diameter ≤2 cm.

Extending 2-D Convolutional Neural Networks to 3-D for Advancing Deep Learning Cancer Classification With Application to MRI Liver Tumor Differentiation

Deep learning (DL) architectures have opened new horizons in medical image analysis attaining unprecedented performance in tasks such as tissue classification and segmentation as well as prediction of several clinical outcomes. In this paper, we propose and evaluate a novel three-dimensional (3-D) convolutional neural network (CNN) designed for tissue classification in medical imaging and applied for discriminating between primary and metastatic liver tumors from diffusion weighted MRI (DW-MRI) data. The proposed network consists of four consecutive strided 3-D convolutional layers with 3 × 3 × 3 kernel size and rectified linear unit (ReLU) as activation function, followed by a fully connected layer with 2048 neurons and a Softmax layer for binary classification. A dataset comprising 130 DW-MRI scans was used for the training and validation of the network. To the best of our knowledge this is the first DL solution for the specific clinical problem and the first 3-D CNN for cancer classification operating directly on whole 3-D tomographic data without the need of any preprocessing step such as region cropping, annotating, or detecting regions of interest. The classification performance results, 83% (3-D) versus 69.6% and 65.2% (2-D), demonstrated significant tissue classification accuracy improvement compared to two 2-D CNNs of different architectures also designed for the specific clinical problem with the same dataset. These results suggest that the proposed 3-D CNN architecture can bring significant benefit in DW-MRI liver discrimination and potentially, in numerous other tissue classification problems based on tomographic data, especially in size-limited, disease-specific clinical datasets.

State of the art in magnetic resonance imaging of hepatocellular carcinoma

Background

Liver cancer is the sixth most common cancer worldwide and the second leading cause of cancer mortality. Chronic liver disease caused by viral infection, alcohol abuse, or other factors can lead to cirrhosis. Cirrhosis is the most important clinical risk factor for hepatocellular carcinoma (HCC) whereby the normal hepatic architecture is replaced by fibrous septa and a spectrum of nodules ranging from benign regenerative nodules to HCC, each one of them with different imaging features.

Conclusions

Furthermore, advanced techniques including the quantification of hepatic and intralesional fat and iron, magnetic resonance elastography, radiomics, radiogenomics, and positron emission tomography (PET)-MRI are highly promising for the extraction of new imaging biomarkers that reflect the tumor microenvironment and, in the future, may add decision-making value in the management of patients with HCC.

Diffusion-weighted imaging in hemorrhagic abdominal and pelvic lesions: restricted diffusion can mimic malignancy

Diffusion-weighted imaging (DWI) is an increasingly utilized sequence in the assessment of abdominal and pelvic lesions. Benign lesions containing hemorrhagic products, with conglomerates of tightly packed blood cells or fibers, can have restricted water diffusion on DWI and apparent diffusion coefficient maps. Such lesions can have restricted diffusion erroneously attributed to malignancy. This review illustrates benign hemorrhagic lesions displaying restricted diffusion, with histopathologic correlation in relevant cases.

Subclinical focal Cholangitis mimicking liver metastasis in asymptomatic patients with history of pancreatic Ductal Adenocarcinoma and Biliary tree intervention

Background

Cholangitis is an inflammatory process of the biliary tract with a wide range of clinical manifestations and it is not always considered in the differential diagnosis in asymptomatic patients. To the best of our knowledge there is no previous report in the English literature of focal cholangitis manifesting exclusively as liver parenchymal changes mimicking liver metastasis in asymptomatic patients with pancreatic ductal adenocarcinoma (PDAC) and history of manipulation of the biliary tree. The purpose of this article is to present six cases of subclinical focal cholangitis mimicking liver metastasis in asymptomatic patients with history of PDAC and biliary tree intervention.

Case presentation

There are six cases with new hepatic lesions detected on follow-up scans in asymptomatic patients with history of PDAC and manipulation of biliary tree. Overall seven lesions were detected, all of them were on the liver periphery, five were hypovascular and two were hypervascular. None of those patients had elevation of CA 19.9 compared with the previous exams. The three patients that had magnetic resonance imaging presented restriction on diffusion weighted imaging and high signal intensity on T2-weighted image. Two patients underwent liver biopsy, which showed only inflammatory changes. All patients were treated with antibiotics and underwent imaging follow-up, which demonstrated resolution of the lesions. None of the patients showed imaging or clinical signs of disease progression during this interval.

Conclusion

Radiologists and oncologists need to be aware of the possibility of focal cholangitis causing hepatic lesions mimicking neoplasia in patients with history of biliary tree intervention, even in the absence of clinical symptoms.

High-flow haemangiomas versus hypervascular hepatocellular carcinoma showing "pseudo-washout" on gadoxetic acid-enhanced hepatic MRI: value of diffusion-weighted imaging in the differential diagnosis of small lesions

AIM

To validate the usefulness of diffusion-weighted imaging (DWI) in the differentiation of high-flow haemangiomas showing pseudo-washout appearance on gadoxetic acid-enhanced hepatic MRI from small hypervascular hepatocellular carcinomas (HCCs).

MATERIALS AND METHODS

DWI (b=50, 800 s/mm²) with apparent diffusion coefficient (ADC) maps for 50 haemangiomas (6.4±2.9 mm) showing intense enhancement on arterial dominant phase imaging and hypointensity on transitional and/or hepatobiliary phase imaging during gadoxetic acid-enhanced MRI were retrospectively analysed and compared with that of 113 hypervascular HCCs (12.8±3.7 mm). In addition to measurement of mean ADC values on DWI and contrast-to-noise ratio (CNR) on corresponding T2-weighted imaging, qualitative analysis of DWI was performed for each lesion by two independent observers using a five-point scale.

RESULTS

Both of mean ADC value (1.902 versus 0.997×10^-3 mm²/s) and mean CNR (119.2 versus 36.9) for haemangioma were significantly larger than for HCC (p<0.001). On receiver operating characteristic (ROC) analysis, an area under the curve (AUC) of 0.995 for ADC values was significantly larger than 0.915 for CNRs (p=0.002). When the ADC value of 1.327×10^-3 mm²/s was used as the threshold for the diagnosis of haemangioma, the sensitivity and specificity were 98% and 97.3%, respectively. The mean sensitivity and specificity of qualitative analysis for the differentiation of haemangioma from HCC were 92% and 99.1%, respectively.

CONCLUSION

For high-flow small haemangiomas showing pseudo-washout appearance during gadoxetic acid-enhanced hepatic MRI, high b-factor DWI including an ADC map may provide additional information to enhance the confidence to exclude small hypervascular HCCs.

Diffusion weighted imaging: Technique and applications

Diffusion weighted imaging (DWI) is a method of signal contrast generation based on the differences in Brownian motion. DWI is a method to evaluate the molecular function and micro-architecture of the human body. DWI signal contrast can be quantified by apparent diffusion coefficient maps and it acts as a tool for treatment response evaluation and assessment of disease progression. Ability to detect and quantify the anisotropy of diffusion leads to a new paradigm called diffusion tensor imaging (DTI). DTI is a tool for assessment of the organs with highly organised fibre structure. DWI forms an integral part of modern state-of-art magnetic resonance imaging and is indispensable in neuroimaging and oncology. DWI is a field that has been undergoing rapid technical evolution and its applications are increasing every day. This review article provides insights in to the evolution of DWI as a new imaging paradigm and provides a summary of current role of DWI in various disease processes.

The Fontan circulation and the liver: A magnetic resonance diffusion-weighted imaging study

BACKGROUND

Patients with a Fontan circulation tend to develop liver fibrosis, liver cirrhosis and even hepatocellular carcinoma. The aim of this study is to use the magnetic resonance technique diffusing-weighted imaging (DWI) for detecting liver fibrosis/cirrhosis in Fontan patients and to establish whether DWI results are associated with functional aspects of the Fontan circulation.

METHODS

In a cross-sectional study, 59 Fontan patients were evaluated by liver DWI. The association between apparent diffusion coefficients (ADC) and patient characteristics, laboratory measurements and functional aspects of the Fontan circulation (NYHA class, maximum oxygen uptake during exercise and cardiac index) was assessed.

RESULTS

Liver ADC values were low (0.82×10(-3)±0.11×10(-3) mm2/s) compared with literature values for healthy volunteers and correlated negatively with calculated liver fibrosis/cirrhosis scores (Fib-4 score, p=0.019; AST/ALT ratio, p=0.009) and gamma-glutamyl transferase (p=0.001). Furthermore, ADC values correlated negatively with follow-up duration (p<0.001) and positively with cardiac index (p=0.019). No correlation between ADC values and exercise tests was found. In multivariable analysis, the ADC values were independently correlated with follow-up duration after Fontan completion.

CONCLUSIONS

The results of the current study suggest that progressive liver damage due to chronic congestion and potential hypoperfusion is reflected in the liver ADC values in Fontan patients. This study highlights that liver damage in the context of the Fontan circulation might be far more common than previously thought, and that the implementation of liver assessment in the routine follow-up of Fontan patients is recommendable.

Lesion discrimination with breath-hold hepatic diffusion-weighted imaging: A meta-analysis

AIM: To investigate the diagnostic capability of breath-hold diffusion-weighted imaging (DWI) for differentiation between malignant and benign hepatic lesions.

METHODS: A total of 614 malignant liver lesions (132 hepatocellular carcinomas, 468 metastases and 14 intrahepatic cholangiocarcinomas) and 291 benign liver lesions (102 hemangiomas, 158 cysts, 24 focal nodular hyperplasia, 1 angiomyolipoma and 6 hepatic adenomas) were included from seven studies (eight sets of data).

RESULTS: The pooled sensitivity and specificity of breath-hold DWI were 0.93 [95% confidence interval (CI): 0.91-0.95] and 0.87 (95%CI: 0.83-0.91), respectively. The positive likelihood ratio and negative likelihood ratio were 7.28 (95%CI: 4.51-11.76) and 0.09 (95%CI: 0.05-0.17), respectively. The P value for χ² heterogeneity for all pooled estimates was < 0.05. From the fitted summary receiver operating characteristic curve, the area under the curve and Q* index were 0.96 and 0.91, respectively. Publication bias was not present (t = 0.49, P = 0.64). The meta-regression analysis indicated that evaluated covariates including magnetic resonance imaging modality, echo time, mean age, maximum b factor, and number of b factors were not sources of heterogeneity (all P > 0.05).

CONCLUSION: Breath-hold DWI is useful for differentiating between malignant and benign hepatic lesions. The diffusion characteristics of benign lesions that mimic malignant ones have rarely been investigated.

Differential diagnosis between hepatic metastases and benign focal lesions using DWI with parallel acquisition technique: a meta-analysis

We aim to investigate the diagnostic capability of diffusion-weighted imaging using parallel acquisition technique for the differentiation between hepatic metastases and benign focal lesions with a meta-analysis. The meta-analysis included a total of 858 hepatic metastases and 440 benign liver lesions from nine studies. The pooled sensitivity and specificity of diffusion-weighted imaging (DWI) were 0.87 (95% CI, 0.84-0.89) and 0.90 (95% CI, 0.87-0.93), respectively. The positive likelihood ratio (PLR) and negative likelihood ratio (NLR) were 8.50 (95% CI, 4.97-14.52) and 0.17 (95% CI, 0.11-0.26), respectively. The P value for χ (2) heterogeneity for all pooled estimates was <0.05. From the fitted summary receiver operating characteristics (SROC), the area under the curve (AUC) and Q* index were 0.95 and 0.88, respectively. Publication bias is not present (t = -0.76, P = 0.471). The meta-regression analysis indicated that evaluated covariates included patient number, patient population, mean age, maximum of b factor, number of cysts, number of hemangiomas, and field were not sources of heterogeneity (all P value >0.05). Diffusion-weighted imaging was useful for differentiation between hepatic metastases and benign focal lesions. The diffusion characteristics of the benign hepatocellular lesions, including cases of focal nodular hyperplasia (FNH) and adenoma, have rarely been reported and need further studies. The diagnostic capability of DWI with parallel acquisition technique for differentiation between metastases and benign hepatic focal lesions might be overestimated.