Digital image processing is one of the most widely used computer vision technologies in biomedical engineering. In the present modern ophthalmological practice, biomarkers analysis through digital fundus image processing analysis greatly contributes to vision science. This further facilitates developments in medical imaging, enabling this robust technology to attain extensive scopes in biomedical engineering platform. Various diagnostic techniques are used to analyze retinal microvasculature image to enable geometric features measurements such as vessel tortuosity, branching angles, branching coefficient, vessel diameter, and fractal dimension. These extracted markers or characterized fundus digital image features provide insights and relates quantitative retinal vascular topography abnormalities to various pathologies such as diabetic retinopathy, macular degeneration, hypertensive retinopathy, transient ischemic attack, neovascular glaucoma, and cardiovascular diseases. Apart from that, this noninvasive research tool is automated, allowing it to be used in large-scale screening programs, and all are described in this present review paper. This paper will also review recent research on the image processing-based extraction techniques of the quantitative retinal microvascular feature. It mainly focuses on features associated with the early symptom of transient ischemic attack or sharp stroke.

Image recognition and understanding is considered as a remarkable subfield of Artificial Intelligence (AI). In practice, retinal image data have high dimensionality leading to enormous size data. As the morphological retinal image datasets can be analyzed in an expansive and non-invasive way, AI more precisely Deep Learning (DL) methods are facilitating in developing intelligent retinal image analysis tools. The most recently developed DL technique, Convolutional Neural Network (CNN) showed remarkable efficiency in identifying, localizing, and quantifying the complex and hierarchical image features that are responsible for severe cardiovascular diseases. Different deep layered CNN architectures such as LeeNet, AlexNet, and ResNet have been developed exploiting CNN morphology. This wide variety of CNN structures can iteratively learn complex data structures of different datasets through supervised or unsupervised learning and perform exquisite analysis for feature recognition independently to diagnose threatening cardiovascular diseases. In modern ophthalmic practice, DL based automated methods are being used in retinopathy screening, grading, identifying, and quantifying the pathological features to employ further therapeutic approaches and offering a wide potentiality to get rid of ophthalmic system complexity. In this review, the recent advances of DL technologies in retinal image segmentation and feature extraction are extensively discussed. To accomplish this study the pertinent materials were extracted from different publicly available databases and online sources deploying the relevant keywords that includes retinal imaging, artificial intelligence, deep learning and retinal database. For the associated publications the reference lists of selected articles were further investigated.

No abstract

Digital image processing is one of the most widely used computer vision techniques, especially in biomedical engineering. Modern ophthalmology is directly dependent on this robust technology, digital image processing to find out the biomarkers analyzing the fundus eye images that are responsible for different kinds of life-threatening diseases like hypertensive retinopathy, Transient Ischemic Attack or sharp stroke. The geometric features like vessel tortuosity, branching angles, vessel diameter, and fractal dimension are considered as the biomarkers for the abovementioned cardiovascular diseases. Retinal vessel diameter widening has found as the early symptom of transient ischemic attack or sharp stroke. In this paper, a completely new and computer-aided automated method to measure the retinal vessel diameter by employing the Euclidean Distance Transform technique was developed. The proposed algorithm measures the Euclidean Distance of the bright pixels exist on the Region of Interest (ROI). Further, the Vascular Disease Image Set (VDIS) and Central Light Reflex Image Set (CLRIS) of Retinal Vessel Image Set for Estimation of Width database were used to evaluate the performance of the proposed algorithm that measures the vessel diameter. The proposed algorithm obtained 98.1% accuracy for the CLRIS and 97.7% accuracy for VDIS. With further evaluation, validation and enhancement of the method, it can be integrated into the clinical computer-aided diagnostic tool.

Diabetic Retinopathy (DR) grading into different stages of severity continues to remain a challenging issue due to the complexities of the disease. Diabetic Retinopathy grading classifies retinal images to five levels of severity ranging from 0 to 5, which represents No DR, Mild non-proliferative diabetic retinopathy (NPDR), Moderate NPDR, Severe NPDR, and proliferative diabetic retinopathy. With the advancement of Deep Learning, studies on the application of the Convolutional Neural Network (CNN) in DR grading have been on the rise. High accuracy and sensitivity are the desired outcome of these studies. This paper reviewed recently published studies that employed CNN for DR grading to 5 levels of severity. Various approaches are applied in classifying retinal images which are, (i) by training CNN models to learn the features for each grade and (ii) by detecting and segmenting lesions using information about their location such as microaneurysms, exudates, and haemorrhages. Public and private datasets have been utilised by researchers in classifying retinal images for DR. The performance of the CNN models was measured by accuracy, specificity, sensitivity, and area under the curve. The CNN models and their performance varies for every study. More research into the CNN model is necessary for future work to improve model performance in DR grading. The Inception model can be used as a starting point for subsequent research. It will also be necessary to investigate the attributes that the model uses for grading.