China 
Africa 
Explore chapters and articles related to this topic
Preliminary Study of Retinal Lesions Classification on Retinal Fundus Images for the Diagnosis of Retinal Diseases
Published in Mitul Kumar Ahirwal, Narendra D. Londhe, Anil Kumar, Artificial Intelligence Applications for Health Care, 2022
Jaskirat Kaur, Ramanpreet Kaur, Deepti Mittal
Cotton wool spots (or soft exudates) are small, swollen micro infarcts that appear because of obstructed blood vessels resulting in the impaired blood supply to that area. Furthermore, the decreased blood flow injures the nerve fibres in that location resulting in blood circulation in local capillaries. They appear on fundus images as yellowish or white fluffy patches with blurred edges as depicted in Figure 7.5. Usually, their local spread is less than 1/3 optic disk areas in diameter and they are generally found throughout the retina. Cotton wool spots alone do not induce vision problems but are strongly associated with disorders influencing the development off the eye, as in the case of systemic retinal diseases. They also typically appear with other retinal abnormalities that cause significant symptoms and have long-term implications. Cotton wool spots sometimes vanish alone although any scattered lack of vision can be irreversible.
Studies on Registration and Fusion of Retinal Images
Published in Rick S. Blum, Zheng Liu, Multi-Sensor Image Fusion and Its Applications, 2018
France Laliberté, Langis Gagnon
The eye (Figure 3.1) can be imaged under many modalities with various sensors. We focus on the imaging conditions of the eye fundus (Figure 3.2). There are two main diseases associated to the retina and choroid: diabetic retinopathy and age-related macular degeneration. Diabetes can cause a weakening of blood vessel bodies, in particular within the retina. The capillaries can leak and become potential hosts for microaneurisms (small bulges that develop in the capillary walls and are the first indicator of diabetic retinopathy). The development of new weak capillaries (neovascularization), which leak easily, can also occur. An edema appears when fluids accumulate in the retina. These fluids are also responsible for exudates, which are metabolic waste products. Closure of capillaries is another possible change that may lead to a lack of oxygen in the retina (ischemia). The cotton wool spots indicate the areas of oxygen-starved retina. Figure 3.3 shows the fundus of a normal eye with the principal anatomical components identified. Figure 3.4 shows the fundus of an eye with diabetic retinopathy on which some common lesions are identified. Choroidal neovascularization associated with age-related macular degeneration leads to irreversible damage in retinal tissue. Early and complete photocoagulation of the affected areas is the only treatment for delaying or preventing decreases in visual acuity.
An Improved Method for Automated Identification of Hard Exudates in Diabetic Retinopathy Disease
Published in IETE Journal of Research, 2022
Niladri Sekhar Datta, Himadri Sekhar Dutta, Koushik Majumder, Sumana Chatterjee, Najir Abdul Wasim
Resource and time are the two main constraints for the manual screening of ocular diseases like Diabetic Retinopathy (DR) and Glaucoma. Presently, to overcome these, there is an emerging trend to set up automated medical diagnosis systems that can efficiently screen a number of patients in time. Clinically, DR, the progressive ocular disease identification at an early stage may reduce the adult blindness by 90% [1]. Globally, 285 million people are affected by DR and the ratio of patient and Ophthalmologists is very high [2]. So, the lack of ophthalmologists increases the need of DR screening system [1,2]. Pathological symptoms on the retina for DR are Microaneurysms, Hemorrhages, Hard Exudates and Cotton-wool spots [3,4]. Microaneurysms and Hemorrhages appear as red lesions in DR-affected retinal images. But, Hard Exudates and Cotton-wool spots may be visualized as yellowish lesions. In DR, retinal blood vessels are damaged and fat-protein based elements are leaked out from the damaged blood vessels. It causes reduction of retinal size. These elements are clinically recognized as Hard Exudates [2]. In time detection of these, the occurrence of blindness for diabetic patients is surely reduced. In this effect, accurate Hard Exudate identification through automated screening depends on the proper pre-processing of the retinal fundus images. Hence, improvement of pre-processing task is an open research problem. The quality of retinal fundus images can affect directly the accuracy of Hard Exudate screening system. So, in automated screening, contrast enhancement of input image is a significant issue. These specific steps of image processing find out the finer details of the object. In our previous work, we presented a highly effective pre-processing method for microaneurysm detection and showed the importance of brightness preservation in medical image analysis [1]. A specific pre-processing method on the retinal image is required for the identification of specific pathological sign in automated screening system, as reported in [1]. The current research is the extension of our previous work. Here, a new pre-processing stage is proposed which is very effective for identifications of Hard Exudates. The fuzzy rules for the digital image are used here and as a result, the intensity values are better handled for contrast enhancement of retinal images. Smooth histogram without missing intensity level is obtained from pre-processed images. Testing result shows that the pre-processing stage produces the enhanced quality retinal images and capable of preserving the mean brightness efficiently. This pre-processing scheme is applied for the Hard Exudate detection method and the testing result indicates the significant improvement of the overall screening system. The organization of the paper is stated as follows. In Section 2, the literature survey is provided. In Section 3, the pre-processing scheme is represented. Hard Exudate identification is presented in Section 4. In Section 5, Results and Discussions of this experiment are given. Finally, the paper is concluded in Section 6.