Hepatoid adenocarcinoma (HAC) of the pancreas is a rare, highly malignant tumor with poor prognosis. This article presents the CT and MR features of HAC of the pancreas, while also reviewing the relevant literatures to succinctly summarize the underlying pathophysiological mechanism, imaging diagnosis, and differential diagnosis. The objective is to enhance the understanding of HAC of the pancreas among clinicians and radiologists.
ObjectiveTo systematically review the value of radiomics in the diagnosis of glioblastoma. MethodsPubMed, EMbase, Web of Science and The Cochrane Library databases were electronically searched to collect studies on radiomics in the grading of gliomas or the differentiation diagnosis from inception to May 30th, 2021. Two reviewers independently screened literature, extracted data, and assessed the risk of bias and the quality of the included studies. Meta-analysis was then performed using Meta-Disc 1.4 software and RevMan 5.3 software. ResultsA total of 37 studies involving 2 746 subjects were included. The results of meta-analysis showed that the pooled sensitivity, specificity, and diagnostic odds ratio for the diagnosis of glioblastoma by radiomics were 0.91 (95%CI 0.89 to 0.92), 0.88 (95%CI 0.87 to 0.90), and 78.00 (95%CI 50.81 to 119.72), respectively. The area under the summary receiver operating characteristic (SROC) curve was 0.95. The key radiomic features for correct diagnosis of glioblastoma included intensity features and texture features of the lesions. ConclusionThe current evidence shows that radiomics provides good diagnostic accuracy for glioblastoma. Due to the limited quality and quantity of the included studies, more high-quality studies are required to verify the above conclusions.