The retinal vessel is the only microvascular network that can be directly and non-invasively observed in humans. Cardiovascular and cerebrovascular diseases, such as diabetes, hypertension, can lead to structural changes of the retinal microvascular network. Therefore, it is of great significance to study effective retinal vessel segmentation methods and assist doctors in early diagnoses with quantitative results for vascular networks. In this study, we propose a novel convolutional neural network named feature pyramid U-Net (FPU-Net) that extracts multiscale representations by constructing two feature pyramids both on the encoder and the decoder of U-Net. In this representation, objects features with different size like micro-vessels and pathology will be fused for better vessel segmentation. The experimental results show that compared with state-of-the-art methods, FPU-Net is superior in terms of accuracy, sensitivity, F1-score, and area under the curve and capable of stronger domain generalisation across different datasets.
INTRODUCTIONMedical image analysis is essential to diagnose the various types of diseases. Segmentation is an important step in many medical applications involving measurements, registration, visualisation, and computer-aided diagnosis (CAD). To help clinicians make an accurate diagnosis, it is necessary to segment some crucial objects in medical images and extract features from segmented areas to make anatomical or pathological structures changes more clear in images and play a pivotal role in CAD on account of the improvement in diagnostic efficiency and accuracy [1].The eye is one of the most important organs through which human receives external information. Approximately 80% of the information received by a person comes from visual perception. Most ophthalmic diseases are caused by fundus retinopathy [2], and the colour fundus image is the one crucial tool that can capture the human microvascular network directly and non-invasively. Doctors can judge the severity of related diseases by observing the structural characteristics of the retinal This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited.