Search

Diabetic retinopathy is one of the main causes of blindness in human eyes, and lesion segmentation is an important basic work for the diagnosis of diabetic retinopathy. Due to the small lesion areas scattered in fundus images, it is laborious to segment the lesion of diabetic retinopathy effectively with the existing U-Net model. In this paper, we proposed a new lesion segmentation model named FFU-Net (Feature Fusion U-Net) that enhances U-Net from the following points. Firstly, the pooling layer in the network is replaced with a convolutional layer to reduce spatial loss of the fundus image. Then, we integrate multiscale feature fusion (MSFF) block into the encoders which helps the network to learn multiscale features efficiently and enrich the information carried with skip connection and lower-resolution decoder by fusing contextual channel attention (CCA) models. Finally, in order to solve the problems of data imbalance and misclassification, we present a Balanced Focal Loss function. In the experiments on benchmark dataset IDRID, we make an ablation study to verify the effectiveness of each component and compare FFU-Net against several state-of-the-art models. In comparison with baseline U-Net, FFU-Net improves the segmentation performance by 11.97%, 10.68%, and 5.79% on metrics SEN, IOU, and DICE, respectively. The quantitative and qualitative results demonstrate the superiority of our FFU-Net in the task of lesion segmentation of diabetic retinopathy. [ABSTRACT FROM AUTHOR]

Two subtypes of influenza A virus (IAV), avian-origin canine influenza virus (CIV) H3N2 (CIV-H3N2) and equine-origin CIV H3N8 (CIV-H3N8), are enzootic in the canine population. Dogs have been demonstrated to seroconvert in response to diverse IAVs, and naturally occurring reassortants of CIV-H3N2 and the 2009 H1N1 pandemic virus (pdmH1N1) have been isolated. We conducted a thorough phenotypic evaluation of CIV-H3N2 in order to assess its threat to human health. Using ferret-generated antiserum, we determined that CIV-H3N2 is antigenically distinct from contemporary human H3N2 IAVs, suggesting that there may be minimal herd immunity in humans. We assessed the public health risk of CIV-H3N2 × pandemic H1N1 (pdmH1N1) reassortants by characterizing their in vitro genetic compatibility and in vivo pathogenicity and transmissibility. Using a luciferase minigenome assay, we quantified the polymerase activity of all possible 16 ribonucleoprotein (RNP) complexes (PB2, PB1, PA, NP) between CIV-H3N2 and pdmH1N1, identifying some combinations that were more active than either parental virus complex. Using reverse genetics and fixing the CIV-H3N2 hemagglutinin (HA), we found that 51 of the 127 possible reassortant viruses were viable and able to be rescued. Nineteen of these reassortant viruses had high-growth phenotypes in vitro, and 13 of these replicated in mouse lungs. A single reassortant with the NP and HA gene segments from CIV-H3N2 was selected for characterization in ferrets. The reassortant was efficiently transmitted by contact but not by the airborne route and was pathogenic in ferrets. Our results suggest that CIV-H3N2 reassortants may pose a moderate risk to public health and that the canine host should be monitored for emerging IAVs. [ABSTRACT FROM AUTHOR]