Cardiovascular resonance (CMR) imaging is a standard imaging modality for assessing cardiovascular diseases (CVDs), the leading cause of death globally. CMR enables accurate quantification of the cardiac chamber volume, ejection fraction and myocardial mass, providing information for diagnosis and monitoring of CVDs. However, for years, clinicians have been relying on manual approaches for CMR image analysis, which is time consuming and prone to subjective errors. It is a major clinical challenge to automatically derive quantitative and clinically relevant information from CMR images. Deep neural networks have shown a great potential in image pattern recognition and segmentation for a variety of tasks. Here we demonstrate an automated analysis method for CMR images, which is based on a fully convolutional network (FCN). The network is trained and evaluated on a large-scale dataset from the UK Biobank, consisting of 4,875 subjects with 93,500 pixelwise annotated images. The performance of the method has been evaluated using a number of technical metrics, including the Dice metric, mean contour distance and Hausdorff distance, as well as clinically relevant measures, including left ventricle (LV) end-diastolic volume (LVEDV) and end-systolic volume (LVESV), LV mass (LVM); right ventricle (RV) end-diastolic volume (RVEDV) and end-systolic volume (RVESV). By combining FCN with a large-scale annotated dataset, the proposed automated method achieves a high performance in segmenting the LV and RV on short-axis CMR images and the left atrium (LA) and right atrium (RA) on long-axis CMR images. On a short-axis image test set of 600 subjects, it achieves an average Dice metric of 0.94 for the LV cavity, 0.88 for the LV myocardium and 0.90 for the RV cavity. The mean absolute difference between automated measurement and manual measurement is 6.1 mL for LVEDV, 5.3 mL for LVESV, 6.9 gram for LVM, 8.5 mL for RVEDV and 7.2 mL for RVESV. On long-axis image test sets, the average Dice metric is 0.93 for the LA cavity (2-chamber view), 0.95 for the LA cavity (4-chamber view) and 0.96 for the RA cavity (4-chamber view). The performance is comparable to human inter-observer variability. We show that an automated method achieves a performance on par with human experts in analysing CMR images and deriving clinically relevant measures.

Purpose: Although microRNAs have recently been recognized as riboregulators of gene expression, little is known about microRNA expression profiles in serous ovarian carcinoma. We assessed the expression of microRNA and the association between microRNA expression and the prognosis of serous ovarian carcinoma. Experimental Design: Twenty patients diagnosed with serous ovarian carcinoma and eight patients treated for benign uterine disease between December 2000 and September 2003 were enrolled in this study. The microRNA expression profiles were examined using DNA microarray and Northern blot analyses. Results: Several microRNAs were differentially expressed in serous ovarian carcinoma compared with normal ovarian tissues, including miR-21, miR-125a, miR-125b, miR-100, miR-145, miR-16, and miR-99a, which were each differentially expressed in >16 patients. In addition, the expression levels of some microRNAs were correlated with the survival in patients with serous ovarian carcinoma. Higher expression of miR-200, miR-141, miR-18a, miR-93, and miR-429, and lower expression of let-7b, and miR-199a were significantly correlated with a poor prognosis (P < 0.05). Conclusion: Our results indicate that dysregulation of microRNAs is involved in ovarian carcinogenesis and associated with the prognosis of serous ovarian carcinoma.

All cancers harbor molecular alterations in their genomes. The transcriptional consequences of these somatic mutations have not yet been comprehensively explored in lung cancer. Here we present the first large scale RNA sequencing study of lung adenocarcinoma, demonstrating its power to identify somatic point mutations as well as transcriptional variants such as gene fusions, alternative splicing events, and expression outliers. Our results reveal the genetic basis of 200 lung adenocarcinomas in Koreans including deep characterization of 87 surgical specimens by transcriptome sequencing. We identified driver somatic mutations in cancer genes including EGFR, KRAS, NRAS, BRAF, PIK3CA, MET, and CTNNB1. Candidates for novel driver mutations were also identified in genes newly implicated in lung adenocarcinoma such as LMTK2, ARID1A, NOTCH2, and SMARCA4. We found 45 fusion genes, eight of which were chimeric tyrosine kinases involving ALK, RET, ROS1, FGFR2, AXL, and PDGFRA. Among 17 recurrent alternative splicing events, we identified exon 14 skipping in the proto-oncogene MET as highly likely to be a cancer driver. The number of somatic mutations and expression outliers varied markedly between individual cancers and was strongly correlated with smoking history of patients. We identified genomic blocks within which gene expression levels were consistently increased or decreased that could be explained by copy number alterations in samples. We also found an association between lymph node metastasis and somatic mutations in TP53. These findings broaden our understanding of lung adenocarcinoma and may also lead to new diagnostic and therapeutic approaches.

The upper respiratory tract is compromised in the early period of COVID-19, but SARS-CoV-2 tropism at the cellular level is not fully defined. Unlike recent single-cell RNA-Seq analyses indicating uniformly low mRNA expression of SARS-CoV-2 entry–related host molecules in all nasal epithelial cells, we show that the protein levels are relatively high and that their localizations are restricted to the apical side of multiciliated epithelial cells. In addition, we provide evidence in patients with COVID-19 that SARS-CoV-2 is massively detected and replicated within the multiciliated cells. We observed these findings during the early stage of COVID-19, when infected ciliated cells were rapidly replaced by differentiating precursor cells. Moreover, our analyses revealed that SARS-CoV-2 cellular tropism was restricted to the nasal ciliated versus oral squamous epithelium. These results imply that targeting ciliated cells of the nasal epithelium during the early stage of COVID-19 could be an ideal strategy to prevent SARS-CoV-2 propagation.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2), which is the cause of a present pandemic, infects human lung alveolar type 2 (hAT2) cells. Characterizing pathogenesis is crucial for developing vaccines and therapeutics. However, the lack of models mirroring the cellular physiology and pathology of hAT2 cells limits the study. Here, we develop a feeder-free, long-term, three-dimensional (3D) culture technique for hAT2 cells derived from primary human lung tissue and investigate infection response to SARS-CoV-2. By imaging-based analysis and single-cell transcriptome profiling, we reveal rapid viral replication and the increased expression of interferon-associated genes and proinflammatory genes in infected hAT2 cells, indicating a robust endogenous innate immune response. Further tracing of viral mutations acquired during transmission identifies full infection of individual cells effectively from a single viral entry. Our study provides deep insights into the pathogenesis of SARS-CoV-2 and the application of defined 3D hAT2 cultures as models for respiratory diseases.