Background: Age-related macular degeneration (AMD) is one of the leading causes of vision loss in the elderly population. The application of artificial intelligence (AI) provides convenience for the diagnosis of AMD. This systematic review and meta-analysis aimed to quantify the performance of AI in detecting AMD in fundus photographs. Methods: We searched PubMed, Embase, Web of Science and the Cochrane Library before December 31st, 2020 for studies reporting the application of AI in detecting AMD in color fundus photographs. Then, we pooled the data for analysis. PROSPERO registration number: CRD42020197532. Findings: 19 studies were finally selected for systematic review and 13 of them were included in the quantitative synthesis. All studies adopted human graders as reference standard. The pooled area under the receiver operating characteristic curve (AUROC) was 0.983 (95% confidence interval (CI):0.979À0.987). The pooled sensitivity, specificity, and diagnostic odds ratio (DOR) were 0.88 (95% CI:0.88À0.88), 0.90 (95% CI:0.90À0.91), and 275.27 (95% CI:158.43À478.27), respectively. Threshold analysis was performed and a potential threshold effect was detected among the studies (Spearman correlation coefficient: -0.600, P = 0.030), which was the main cause for the heterogeneity. For studies applying convolutional neural networks in the Age-Related Eye Disease Study database, the pooled AUROC, sensitivity, specificity, and DOR were 0.983 (95% CI:0.978À0.988), 0.88 (95% CI:0.88À0.88), 0.91 (95% CI:0.91À0.91), and 273.14 (95% CI:130.79À570.43), respectively. Interpretation: Our data indicated that AI was able to detect AMD in color fundus photographs. The application of AI-based automatic tools is beneficial for the diagnosis of AMD.

Background There is increasing evidence that the lung is a target organ of diabetes. This study aimed to examine in detail the association between diabetes mellitus and pulmonary function using a national cohort. We also aimed to explore the non-linear association between pulmonary function and blood glucose, insulin resistance, and C-reactive protein (CRP). Methods A total of 30,442 participants from the National Health and Nutrition Examination Survey from the period between 2007 and 2012 were included. The cross-sectional association between diabetes mellitus and pulmonary function was assessed using multiple linear regression. Where there was evidence of non-linearity, we applied a restricted cubic spline with three knots to explore the non-linear association. Partial mediation analysis was performed to evaluate the underlying mechanism. All analyses were weighted to represent the US population and to account for the intricate survey design. Results A total of 8584 people were included in the final study population. We found that diabetes was significantly associated with reduced forced expiratory volume in one second (FEV1) and forced vital capacity. We further found L-shaped associations between hemoglobin A1c (HbA1c) and pulmonary function. There was a negative association between HbA1c and FEV1 in diabetes participants with good glucose control (HbA1c < 7.0%), but not in patients with poor glucose control. A non-linear association was also found with fasting plasma glucose, 2 h-plasma glucose after oral glucose tolerance test, insulin resistance, and CRP. Finally, we found that diabetes duration did not affect pulmonary function, and the deleterious effect of diabetes on pulmonary function was mediated by hyperglycemia, insulin resistance, low-grade chronic inflammation (CRP), and obesity. Conclusions Diabetes mellitus is non-linearly associated with pulmonary function. Our finding of a negative association between HbA1c and FEV1 in diabetes patients with good glucose control but not in patients with poor glucose control indicates that a stricter glycemic target should be applied to diabetic patients to improve pulmonary function. Given, the cross-sectional nature of this research, a longitudinal study is still needed to validate our findings.

Infrared scene simulation has extensive application value in military and civil fields. According to the specific experiment environment, this paper simulation of an infrared scene and according to the experimental data produced infrared texture library. First of all, through the 3Dmax modeling software to establish the target infrared radiation model, and calculate the relevant radiation data of it. Then analyze the radiation characteristics of the relevant materials, the establishment of infrared texture library for each material. And finally, use OGRE engine image rendering technology and GPU programmable pipeline technique to simulate a complex infrared scene close to the real.