• Multiparametric ultrasomics has achieved much better performance in the discrimination of significant fibrosis (≥ F2) than the single modality of conventional radiomics, original radiofrequency and contrast-enhanced micro-flow. • Adaptive boosting, random forest and support vector machine are the optimal algorithms for machine learning.

Purpose and methods: The ability of contrast-enhanced ultrasound (CEUS) to differentiate between hepatocellular carcinoma (HCC) and intrahepatic cholangiocarcinoma (ICC) is still controversial. We reviewed the CEUS imaging of 819 patients (HCC=546, ICC=273) with an established pathological diagnosis. The enhancement patterns of lesions and the diagnostic performance of CEUS were analyzed. Results: Arterial hyperenhancement followed by washout was observed in 92.3% (504/546) of the HCC lesions and 85.7% (234/273) of the ICC lesions on CEUS (p<0.05). Additionally, the ICCs presented contrast washout much earlier than the HCCs, with an average time of 27.5 seconds after injecting the contrast agent compared with 70.1 seconds for the HCCs (p<0.05). Peripheral rim-like enhancement was observed in 68.5% (187/273) of the ICCs, which was significantly more common than that in the HCCs (2.0%, 11/546) (p<0.05). When using arterial hyperenhancement with a washout phase later than 43 seconds after injecting the contrast agent and with no peripheral rim-like enhancement as the diagnostic criteria for HCC ≤5 cm in diameter, the area under the curve was 0.808, with 64.1% sensitivity, 97.4% specificity and 73.6% accuracy. Conclusions: Although ICC may show the typical enhancement pattern of HCC on CEUS, peripheral rim-like enhancement and quick contrast washout show high efficiency in the differentiation of HCC from ICC.

Microvascular invasion (MVI) is rarely diagnosed preoperatively in hepatocellular carcinoma (HCC). The aim of this meta-analysis is to assess the diagnostic power of a non-smooth tumor margin on preoperative imaging for MVI. We performed a literature search using the PubMed, Embase and Cochrane Library databases, and 11 studies were included involving 618 MVI-positive cases and 1030 MVI-negative cases. Considerable heterogeneity was found, and was indicated to be attributable to the mean patient ages in the included studies. In subgroups of studies with a mean patient age older than 60 years and studies with computed tomography (CT) as the imaging method (as opposed to magnetic resonance imaging (MRI)), heterogeneity was low, and the diagnostic odds ratio (DOR) of the single two-dimensional imaging feature for MVI was 21.30 (95% CI [12.52, 36.23]) and 28.78 (95% CI [13.92, 59.36]), respectively; this power was equivalent to or greater than that of certain multivariable-based scoring systems. In conclusion, a non-smooth tumor margin on preoperative imaging is of great value for MVI assessment and should be considered for inclusion in future scoring systems.

Serologic studies are urgently needed to assist in understanding an outbreak of influenza A(H7N9) virus. However, a biosafety level 3 laboratory is required for conventional serologic assays with live lethal virus. We describe a safe pseudovirus–based neutralization assay with preliminary assessment using subtype H7N9–infected samples and controls.

No abstract