ABSTRACT
BACKGROUND: Hepatocellular carcinoma (HCC), theseventh most common cancer worldwide, is characterized by a high mortality rate, advanced diagnosis, and susceptibility to extrahepatic metastasis. Numerous studies have shown that DNA methylation is a crucial factor in epigenetic modifications and regulation of carcinogenesis. METHODS: HCC patient data were sourced from the TCGA dataset as a training set, while GSE116174 was used as an external validation set for verification. Differential methylation and expression analyses were performed on HCC samples with and without extrahepatic metastasis. In the intersecting genes, the relationship between methylation and expression levels of the intersecting genes was analyzed. Genes with a correlation coefficient≥|0.30| and P<0.05 were identified as methylation driver genes. Cox regression analysis was conducted to identify genes associated with HCC prognosis and establish a risk score. Subsequently, a prognostic model was established and validated using Cox regression analysis incorporating the risk score and other clinical factors. Using immunohistochemistry to evaluate the expression of DHX58 and EIF5A2 in HCC tissues with and without extrahepatic metastasis. Immunoinfiltration analysis was performed on the HCC samples using CIBERSORT. RESULTS: Our research identified eight methylation driver genes for HCC extrahepatic metastasis, of which two genes (DHX58 and EIF5A2) were associated with HCC patient prognosis. And the study further constructed and validated the risk score and prognostic model. Immunoinfiltration analysis showed that M0 macrophage abundance was correlated with the prognosis of HCC patients. Immunohistochemistry revealed differences in DHX58 and EIF5A2 expression between HCC tissues with and without extrahepatic metastasis, consistent with our bioinformatics findings.
ABSTRACT
Under-display imaging has recently received considerable attention in both academia and industry. As a variation of this technique, under-display ToF (UD-ToF) cameras enable depth sensing for full-screen devices. However, it also brings problems of image blurring, signal-to-noise ratio and ranging accuracy reduction. To address these issues, we propose a cascaded deep network to improve the quality of UD-ToF depth maps. The network comprises two subnets, with the first using a complex-valued network in raw domain to perform denoising, deblurring and raw measurements enhancement jointly, while the second refining depth maps in depth domain based on the proposed multi-scale depth enhancement block (MSDEB). To enable training, we establish a data acquisition device and construct a real UD-ToF dataset by collecting real paired ToF raw data. Besides, we also build a large-scale synthetic UD-ToF dataset through noise analysis. The quantitative and qualitative evaluation results on public datasets and ours demonstrate that the presented network outperforms state-of-the-art algorithms and can further promote full-screen devices in practical applications.