Medical image segmentation is an important step in medical image analysis. With the rapid development of convolutional neural network in image processing, deep learning has been used for medical image segmentation, such as optic disc segmentation, blood vessel detection, lung segmentation, cell segmentation, etc. Previously, U-net based approaches have been proposed. However, the consecutive pooling and strided convolutional operations lead to the loss of some spatial information. In this paper, we propose a context encoder network (referred to as CE-Net) to capture more high-level information and preserve spatial information for 2D medical image segmentation. CE-Net mainly contains three major components: a feature encoder module, a context extractor and a feature decoder module. We use pretrained ResNet block as the fixed feature extractor. The context extractor module is formed by a newly proposed dense atrous convolution (DAC) block and residual multi-kernel pooling (RMP) block. We applied the proposed CE-Net to different 2D medical image segmentation tasks. Comprehensive results show that the proposed method outperforms the original U-Net method and other state-of-the-art methods for optic disc segmentation, vessel detection, lung segmentation, cell contour segmentation and retinal optical coherence tomography layer segmentation.

Issue: 1-2articleIncreasing strength of new structural materials and longer spans of new footbridges, accompanied with aesthetic requirements for greater slenderness, are resulting in more lively footbridge structures. In the past few years this issue attracted great public attention. The excessive lateral sway motion caused by crowd walking across the infamous Millennium Bridge in London is the prime example of the vibration serviceability problem of footbridges. In principle, consideration of footbridge vibration serviceability requires a characterisation of the vibration source, path and receiver. This paper is the most comprehensive review published to date of about 200 references which deal with these three key issues.\ud \ud The literature survey identified humans as the most important source of vibration for footbridges. However, modelling of the crowd-induced dynamic force is not clearly defined yet, despite some serious attempts to tackle this issue in the last few years.\ud \ud The vibration path is the mass, damping and stiffness of the footbridge. Of these, damping is the most uncertain but extremely important parameter as the resonant behaviour tends to govern vibration serviceability of footbridges.\ud \ud A typical receiver of footbridge vibrations is a pedestrian who is quite often the source of vibrations as well. Many scales for rating the human perception of vibrations have been found in the published literature. However, few are applicable to footbridges because a receiver is not stationary but is actually moving across the vibrating structure.\ud \ud During footbridge vibration, especially under crowd load, it seems that some form of human–structure interaction occurs. The problem of influence of walking people on footbridge vibration properties, such as the natural frequency and damping is not well understood, let alone quantified.\ud \ud Finally, there is not a single national or international design guidance which covers all aspects of the problem comprehensively and some form of their combination with other published information is prudent when designing major footbridge structures. The overdue update of the current codes to reflect the recent research achievements is a great challenge for the next 5–10 years

Optical Coherence Tomography Angiography (OCTA) is a non-invasive imaging technique that has been increasingly used to image the retinal vasculature at capillary level resolution. However, automated segmentation of retinal vessels in OCTA has been under-studied due to various challenges such as low capillary visibility and high vessel complexity, despite its significance in understanding many vision-related diseases. In addition, there is no publicly available OCTA dataset with manually graded vessels for training and validation of segmentation algorithms. To address these issues, for the first time in the field of retinal image analysis we construct a dedicated Retinal OCTA SEgmentation dataset (ROSE), which consists of 229 OCTA images with vessel annotations at either centerline-level or pixel level. This dataset with the source code has been released for public access to assist researchers in the community in undertaking research in related topics. Secondly, we introduce a novel split-based coarse-to-fine vessel segmentation network for OCTA images (OCTA-Net), with the ability to detect thick and thin vessels separately. In the OCTA-Net, a split-based coarse segmentation module is first utilized to produce a preliminary confidence map of vessels, and a split-based refined segmentation module is then used to optimize the shape/contour of the retinal microvasculature. We perform a thorough evaluation of the state-of-the-art vessel segmentation models and our OCTA-Net on the constructed ROSE dataset. The experimental results demonstrate that our OCTA-Net yields better vessel segmentation performance in OCTA than both traditional and other deep learning methods. In addition, we provide a fractal dimension analysis on the segmented microvasculature, and the statistical analysis demonstrates significant differences between the healthy control and Alzheimer’s Disease group. This consolidates that the analysis of retinal microvasculature may offer a new scheme to study various neurodegenerative diseases.