Atrial Septal Defect Detection in Children Based on Ultrasound Video Using Multiple Instances Learning.

Thoracic echocardiography (TTE) can provide sufficient cardiac structure information, evaluate hemodynamics and cardiac function, and is an effective method for atrial septal defect (ASD) examination. This paper aims to study a deep learning method based on cardiac ultrasound video to assist in ASD diagnosis. We chose four standard views in pediatric cardiac ultrasound to identify atrial septal defects; the four standard views were as follows: subcostal sagittal view of the atrium septum (subSAS), apical four-chamber view (A4C), the low parasternal four-chamber view (LPS4C), and parasternal short-axis view of large artery (PSAX). We enlist data from 300 children patients as part of a double-blind experiment for five-fold cross-validation to verify the performance of our model. In addition, data from 30 children patients (15 positives and 15 negatives) are collected for clinician testing and compared to our model test results (these 30 samples do not participate in model training). In our model, we present a block random selection, maximal agreement decision, and frame sampling strategy for training and testing respectively, resNet18 and r3D networks are used to extract the frame features and aggregate them to build a rich video-level representation. We validate our model using our private dataset by five cross-validation. For ASD detection, we achieve 89.33 ± 3.13 AUC, 84.95 ± 3.88 accuracy, 85.70 ± 4.91 sensitivity, 81.51 ± 8.15 specificity, and 81.99 ± 5.30 F1 score. The proposed model is a multiple instances learning-based deep learning model for video atrial septal defect detection which effectively improves ASD detection accuracy when compared to the performances of previous networks and clinical doctors.

Is Chinese Spring Festival a key point for glycemic control of patients with type 2 diabetes mellitus in China?

Objectives: This study aims to explore the long-term trend of fasting blood glucose (FBG) among urban patients with type 2 diabetes mellitus (T2DM) and the impacts of the Chinese Spring Festival on their glycemic control in urban China. Methods: The general information and longitudinal monitoring data of patients with T2DM in Minhang District, Shanghai China from 15 December 2006 to 31 December 2015 were collected. The FBG records were grouped into three periods, namely, the preholiday period (2 months right before the Chinese Spring Festival), the holiday period (from 28 December to 7 January of the lunar calendar year), and the postholiday period (2 months after the Chinese Spring Festival). The Mann-Kendall trend test and Cochran-Armitage trend test were occupied to explore the long-term trend, and paired t-test and chi-square (χ2) test were used to determine the differences in glycemic level and control rate between the preholiday and postholiday periods, respectively. Results: From 2007 to 2015, the glycemic control rate in patients with T2DM showed an upward trend (P < 0.001), and the FBG level showed a decreasing trend (P = 0.048). After the Chinese Spring Festival, the glycemic control rate decreased significantly (P < 0.001), and the FBG level increased significantly (P < 0.001) compared to those during the preholiday period. The incidence of hypoglycemia increased during holidays. Patients who were aged 60-69 years, overweight or obese, with hypertension, with a disease duration of <3 years, or with poor glycemic control in one previous year were more likely to be affected by the holiday. Conclusion: Chinese Spring Festival is a key point for glycemic control of patients with T2DM in China. Intensive holiday-specific diabetic healthcare needs to be further improved, and community-based interventions should be developed and implemented to control the possible holiday effects.

Hybrid sequencing of the Gynostemma pentaphyllum transcriptome provides new insights into gypenoside biosynthesis.

BACKGROUND: Gypenosides are a group of triterpene saponins from Gynostemma pentaphyllum that are the same as or very similar to ginsenosides from the Panax species. Several enzymes involved in ginsenoside biosynthesis have been characterized, which provide important clues for elucidating the gypenoside biosynthetic pathway. We suppose that gypenosides and ginsenosides may have a similar biosynthetic mechanism and that the corresponding enzymes in the two pathways may have considerable similarity in their sequences. To further understand gypenoside biosynthesis, we sequenced the G. pentaphyllum transcriptome with a hybrid sequencing-based strategy and then determined the candidate genes involved in this pathway using phylogenetic tree construction and gene expression analysis. RESULTS: Following the PacBio standard analysis pipeline, 66,046 polished consensus sequences were obtained, while Illumina data were assembled into 140,601 unigenes with Trinity software. Then, these output sequences from the two analytical routes were merged. After removing redundant data with CD-HIT software, a total of 140,157 final unigenes were obtained. After functional annotation, five 2,3-oxidosqualene cyclase genes, 145 cytochrome P450 genes and 254 UDP-glycosyltransferase genes were selected for the screening of genes involved in gypenoside biosynthesis. Using phylogenetic analysis, several genes were divided into the same subfamilies or closely related evolutionary branches with characterized enzymes involved in ginsenoside biosynthesis. Using real-time PCR technology, their expression patterns were investigated in different tissues and at different times after methyl jasmonate induction. Since the genes in the same biosynthetic pathway are generally coexpressed, we speculated that GpOSC1, GpCYP89, and GpUGT35 were the leading candidates for gypenoside biosynthesis. In addition, six GpWRKYs and one GpbHLH might play a possible role in regulating gypenoside biosynthesis. CONCLUSIONS: We developed a hybrid sequencing strategy to obtain longer length transcriptomes with increased accuracy, which will greatly contribute to downstream gene screening and characterization, thus improving our ability to elucidate secondary metabolite biosynthetic pathways. With this strategy, we found several candidate genes that may be involved in gypenoside biosynthesis, which laid an important foundation for the elucidation of this biosynthetic pathway, thus greatly contributing to further research in metabolic regulation, synthetic biology and molecular breeding in this species.