Automated retinal vessel segmentation technology has become an important tool for disease screening and diagnosis in clinical medicine. However, most of the available methods of retinal vessel segmentation still have problems such as poor accuracy and low generalization ability. This is because the symmetrical and asymmetrical patterns between blood vessels are complicated, and the contrast between the vessel and the background is relatively low due to illumination and pathology. Robust vessel segmentation of the retinal image is essential for improving the diagnosis of diseases such as vein occlusions and diabetic retinopathy. Automated retinal vein segmentation remains a challenging task. In this paper, we proposed an automatic retinal vessel segmentation framework using deep fully convolutional neural networks (FCN), which integrate novel methods of data preprocessing, data augmentation, and full convolutional neural networks. It is an end-to-end framework that automatically and efficiently performs retinal vessel segmentation. The framework was evaluated on three publicly available standard datasets, achieving F1 score of 0.8321, 0.8531, and 0.8243, an average accuracy of 0.9706, 0.9777, and 0.9773, and average area under the Receiver Operating Characteristic (ROC) curve of 0.9880, 0.9923 and 0.9917 on the DRIVE, STARE, and CHASE_DB1 datasets, respectively. The experimental results show that our proposed framework achieves state-of-the-art vessel segmentation performance in all three benchmark tests.
BackgroundJapanese encephalitis virus (JEV) is the leading cause of viral encephalitis in Asia. Japanese encephalitis (JE) caused by JEV is characterized by extensive inflammatory cytokine secretion, microglia activation, blood-brain barrier (BBB) breakdown, and neuronal death, all of which contribute to the vicious cycle of inflammatory damage. There are currently no effective treatments for JE. Mesenchymal stem cells (MSCs) have been demonstrated to have a therapeutic effect on many central nervous system (CNS) diseases by regulating inflammation and other mechanisms.MethodsIn vivo, 8- to 10-week-old mice were infected intraperitoneally with JEV and syngeneic bone marrow MSCs were administered through the caudal vein at 1 and 3 days post-infection. The mortality, body weight, and behavior were monitored daily. Brains from each group were harvested at the indicated times for hematoxylin and eosin staining, immunohistochemical observation, flow cytometric analysis, TUNEL staining, Western blot, quantitative real-time polymerase chain reaction, and BBB permeability assays. In vitro, co-culture and mixed culture experiments of MSCs with either microglia or neurons were performed, and then the activation state of microglia and survival rate of neurons were tested 48 h post-infection.ResultsMSC treatment reduced JEV-induced mortality and improved the recovery from JE in our mouse model. The inflammatory response, microglia activation, neuronal damage, BBB destruction, and viral load (VL) were significantly decreased in the MSC-treated group. In co-culture experiments, MSCs reprogrammed M1-to-M2 switching in microglia and improved neuron survival. Additionally, the VL was decreased in Neuro2a cells in the presence of MSCs accompanied by increased expression of interferon-α/β.ConclusionMSC treatment alleviated JEV-induced inflammation and mortality in mice.Electronic supplementary materialThe online version of this article (doi:10.1186/s13287-017-0486-5) contains supplementary material, which is available to authorized users.
a b s t r a c tEvaluation for generalization performance of learning algorithms has been the main thread of machine learning theoretical research. The previous bounds describing the generalization performance of the empirical risk minimization (ERM) algorithm are usually established based on independent and identically distributed (i.i.d.) samples. In this paper we go far beyond this classical framework by establishing the generalization bounds of the ERM algorithm with uniformly ergodic Markov chain (u.e.M.c.) samples. We prove the bounds on the rate of uniform convergence/relative uniform convergence of the ERM algorithm with u.e.M.c. samples, and show that the ERM algorithm with u.e.M.c. samples is consistent. The established theory underlies application of ERM type of learning algorithms.
We develop a dynamic valuation model of the hedge fund seeding business by solving the consumption and portfolio-choice problem for a risk-averse manager who launches a hedge fund through a seeding vehicle. This vehicle, i.e. fees-for-seed swap, specifies that a strategic partner (seeder) provides a critical amount of capital in exchange for participation in the funds revenue. Our results indicate that the new swap not only solves the serious problem of widespread financing constraints for new and early-stage funds (ESFs) managers, but can be highly beneficial to both the manager and the seeder if structured properly.
We consider a risk-averse entrepreneur who invests in a project with idiosyncratic risk. In contrast to the literature, we assume the entrepreneur is unable to get a loan from a bank directly because of the low creditability of the entrepreneur and so an innovative financial contract, named equity-forguarantee swap, is signed among a bank, an insurer, and the entrepreneur.It is shown that the new swap leads to higher leverage, which brings more diversification and tax benefits. The new swap not only solves the problems of financing constraints, but also significantly improves the welfare level of the entrepreneur. The growth of welfare level increases dramatically with risk aversion index and the volatility of idiosyncratic risk.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.