No abstract

No abstract

Entropy images, representing the complexity of original fundus photographs, may strengthen the contrast between diabetic retinopathy (DR) lesions and unaffected areas. The aim of this study is to compare the detection performance for severe DR between original fundus photographs and entropy images by deep learning. A sample of 21,123 interpretable fundus photographs obtained from a publicly available data set was expanded to 33,000 images by rotating and flipping. All photographs were transformed into entropy images using block size 9 and downsized to a standard resolution of 100 × 100 pixels. The stages of DR are classified into 5 grades based on the International Clinical Diabetic Retinopathy Disease Severity Scale: Grade 0 (no DR), Grade 1 (mild nonproliferative DR), Grade 2 (moderate nonproliferative DR), Grade 3 (severe nonproliferative DR), and Grade 4 (proliferative DR). Of these 33,000 photographs, 30,000 images were randomly selected as the training set, and the remaining 3,000 images were used as the testing set. Both the original fundus photographs and the entropy images were used as the inputs of convolutional neural network (CNN), and the results of detecting referable DR (Grades 2–4) as the outputs from the two data sets were compared. The detection accuracy, sensitivity, and specificity of using the original fundus photographs data set were 81.80%, 68.36%, 89.87%, respectively, for the entropy images data set, and the figures significantly increased to 86.10%, 73.24%, and 93.81%, respectively (all p values <0.001). The entropy image quantifies the amount of information in the fundus photograph and efficiently accelerates the generating of feature maps in the CNN. The research results draw the conclusion that transformed entropy imaging of fundus photographs can increase the machinery detection accuracy, sensitivity, and specificity of referable DR for the deep learning-based system.

BackgroundThe multidrug-resistant (MDR) and extensively drug-resistant (XDR) tuberculosis (TB) has emerged as a global threat. Xinjiang is a multi-ethnic region and suffered second highest incidence of TB in China. However, epidemiological information on MDR and XDR TB is scarcely investigated.Methodology/Principal FindingsA prospective study was conducted to analyze the prevalence of MDR and XDR TB and the differences of drug resistance TB between Chinese Han and other nationalities population at Chest Hospital of Xinjiang Uygur Autonomous Region, China. We performed in vitro drug susceptibility testing of Mycobacterium tuberculosis to first- and second-line anti-tuberculosis drugs for all 1893 culture confirmed positive TB cases that were diagnosed between June 2009 and June 2011. Totally 1117 (59.0%, 95% CI, 56.8%–61.2%) clinical isolates were resistant to ≥1 first-line drugs; the prevalence of MDR TB was 13.2% (95% CI, 11.7%–14.7%), of which, 77 (30.8%; 95% CI, 25.0%–36.6%) and 31 (12.8%; 95% CI, 8.6%–17.0%) isolates were pre-XDR and XDR TB respectively. Among the MDR/XDR TB, Chinese Han patients were significantly less likely to be younger with an odds ratio 0.42 for age 20–29 years and 0.52 for age 40–49 years; P trend = 0.004), and Chinese Han patients has a lower prevalence of XDR TB (9.6%) than all the other nationality (14.9%).Conclusions/SignificanceThe burden of drug resistance TB cases is sizeable, which highlights an urgent need to reinforce the control, detection and treatment strategies for drug resistance TB. However, the difference of MDR and XDR TB between Chinese Han and other nationalities was not observed.

BackgroundSusceptibility to tuberculosis is not only determined by Mycobacterium tuberculosis infection, but also by the genetic component of the host. Macrophage receptor with a collagenous structure (MARCO) is essential components required for toll like receptor-signaling in macrophage response to Mycobacterium tuberculosis, which may contribute to tuberculosis risk.Principal FindingsTo specifically investigated whether single nucleotide polymorphisms (SNPs) in MARCO gene are associated with pulmonary tuberculosis in Chinese Han population. By selecting tagging SNPs in MARCO gene, 17 tag SNPs were identified and genotyped in 923 pulmonary tuberculosis patients and 1033 healthy control subjects using a hospital based case-control association study. Single-point and haplotype analysis revealed an association in intron and exon region of MARCO gene. One SNP (rs17009726) was associated with susceptibility to pulmonary tuberculosis, where the carriers of the G allele had a 1.65 fold (95% CI = 1.32–2.05, p corrected = 9.27E–5) increased risk of pulmonary tuberculosis. Haplotype analysis revealed that haplotype GC containing G allele of 17009726 and haplotype TGCC (rs17795618T/A, rs1371562G/T, rs6761637T/C, rs2011839C/T) were also associated with susceptibility to pulmonary tuberculosis (p corrected = 0.0001 and 0.029, respectively).ConclusionsOur study suggested that genetic variants in MARCO gene were associated with pulmonary tuberculosis susceptibility in Chinese Han population, and the findings emphasize the importance of MARCO mediated immune responses in the pathogenesis of tuberculosis.