Automated fundus screening is becoming a significant programme of telemedicine in ophthalmology. Instant quality evaluation of uploaded retinal images could decrease unreliable diagnosis. In this work, we propose fractal dimension of retinal vasculature as an easy, effective and explainable indicator of retinal image quality. The pipeline of our approach is as follows: utilize image pre-processing technique to standardize input retinal images from possibly different sources to a uniform style; then, an improved deep learning empowered vessel segmentation model is employed to extract retinal vessels from the pre-processed images; finally, a box counting module is used to measure the fractal dimension of segmented vessel images. A small fractal threshold (could be a value between 1.45 and 1.50) indicates insufficient image quality. Our approach has been validated on 30,644 images from four public database.

Topological and geometrical analysis of retinal blood vessels could be a cost-effective way to detect various common diseases. Automated vessel segmentation and vascular tree analysis models require powerful generalization capability in clinical applications. In this work, we constructed a novel benchmark RETA with 81 labelled vessel masks aiming to facilitate retinal vessel analysis. A semi-automated coarse-to-fine workflow was proposed for vessel annotation task. During database construction, we strived to control inter-annotator and intra-annotator variability by means of multi-stage annotation and label disambiguation on self-developed dedicated software. In addition to binary vessel masks, we obtained other types of annotations including artery/vein masks, vascular skeletons, bifurcations, trees and abnormalities. Subjective and objective quality validations of the annotated vessel masks demonstrated significantly improved quality over the existing open datasets. Our annotation software is also made publicly available serving the purpose of pixel-level vessel visualization. Researchers could develop vessel segmentation algorithms and evaluate segmentation performance using RETA. Moreover, it might promote the study of cross-modality tubular structure segmentation and analysis.

Tessellation in fundus is not only a visible feature for aged-related and myopic maculopathy but also confuse retinal vessel segmentation. The detection of tessellated images is an inevitable processing in retinal image analysis. In this work, we propose a model using convolutional neural network for detecting tessellated images. The input to the model is pre-processed fundus image, and the output indicate whether this photograph has tessellation or not. A database with 12,000 colour retinal images is collected to evaluate the classification performance. The best tessellation classifier achieves accuracy of 97.73% and AUC value of 0.9659 using pretrained GoogLeNet and transfer learning technique.

Diabetic Retinopathy (DR) is a complication of diabetes that can cause vision impairment and lead to permanent blindness if left undiagnosed. The increasing number of diabetic patients, coupled with a shortage of ophthalmologists, highlights the urgent need for automated screening tools for early DR diagnosis. Among the earliest and most detectable signs of DR are microaneurysms (MAs). However, detecting MAs in fundus images remains challenging due to several factors, including image quality limitations, the subtle appearance of MA features, and the wide variability in color, shape, and texture. To address these challenges, we propose a novel preprocessing pipeline that enhances the overall image quality, facilitating feature learning and improving the detection of subtle MA features in low-quality fundus images. Building on this preprocessing technique, we further develop a lightweight Attention U-Net model that significantly reduces the number of model parameters while achieving superior performance. By incorporating an attention mechanism, the model focuses on the subtle features of MAs, leading to more precise segmentation results. We evaluated our method on the IDRID dataset, achieving a sensitivity of 0.81 and specificity of 0.99, outperforming existing MA segmentation models. To validate its generalizability, we tested it on the E-Ophtha dataset, where it achieved a sensitivity of 0.59 and specificity of 0.99. Despite its lightweight design, our model demonstrates robust performance under challenging conditions such as noise and varying lighting, making it a promising tool for clinical applications and large-scale DR screening.