China
Africa
Explore chapters and articles related to this topic
Detection of Eye Diseases using Image Processing and Artificial Neural Networks
Published in Durgesh Kumar Mishra, Nilanjan Dey, Bharat Singh Deora, Amit Joshi, ICT for Competitive Strategies, 2020
K. Sujatha, V. Srividhya, V. Karthikeyan, L. Madheshwaran, N. P. G. Bhavani
Diabetic retinopathy in Figure 3.2(a), results due to insulin disorders causing diabetes. The blood vessels in the light sensitive region retina are affected. It is because of insufficient supply of oxygen leading to blindness. If this eye disorder is diagnosed at early stage proper treatment can be given preventing blindness. The two major categories are proliferative and non- proliferative retinopathy. Non-proliferative retinopathy is less severe and causes hemorrhage in the retina. This produces a leak in blood serum making the retina wet which leads to diminished vision. The severe type is Proliferative retinopathy which produces new fragile blood vessels on the retina. These vessels frequently bleed into the vitreous, the clear jelly in the center of the eyes causing visual problems. It is treated by laser surgery which will reduce the progression of diabetic retinopathy and at times will reverse visual loss causing permanent damage. If Diabetic retinopathy is identified at early stages a better control of blood sugar can be maintained by ensuring lifestyle modification, including abrupt weight loss, dietary changes and simple exercises [3, 6].
Biomimetic Microsystems for Blood and Lymphatic Vascular Research
Published in Hyun Jung Kim, Biomimetic Microengineering, 2020
Here, so far, we have only discussed a few of the physiological and pathological conditions where angiogenesis contributes. We encourage the readers to inquire a more detailed list of angiogenesis-related diseases elsewhere (Carmeliet 2003). For example, bone fracture fails to heal when angiogenesis inhibitors are used. Serious complication of rheumatoid arthritis can lead to vasculitis where the blood vessels are inflamed and sometimes become narrow to prevent adequate blood flow. In the eye, diabetic retinopathy is a complication in diabetic patients. Blood vessels are also inflamed in diabetic retinopathy and can leak fluid into the eye. At the advanced stage, angiogenesis occurs to cause excessive outgrowth of abnormal blood vessels in the eye. If not properly treated, diabetic retinopathy can lead to permanent vision loss (Carmeliet 2003).
*
Published in P. Dakin John, G. W. Brown Robert, Handbook of Optoelectronics, 2017
Constantinos Pitris, Tuan Vo-Dinh, R. Eugene Goodson, Susie E. Goodson
Laser-induced photocoagulation or ablation can be used to alter the tissue shape for surgical or other therapeutic purposes. It is based on the absorption of high-intensity pulses by the targeted tissues causing either protein denaturation or complete evaporation without carbonizing or bleeding. The precise control of the wavelength as well as temporal and power parameters of laser therapeutic techniques can restrict the interaction to specific target areas of tissue. Laser therapy is the current standard of care for the treatment of some retinal diseases such as proliferative diabetic retinopathy, diabetic macular edema, and some types of subretinal neovascularization [72]. Vision correction using photorefractive keratectomy or laser-assisted in situ keratomileusis is also based on this effect [73]. In dermatology, careful control of laser parameters permits selective destruction of specific loci in the skin, for example, in tattoo removal, treatment of port-wine stains, and various cosmetic applications (Figure 26.6) [74].
Feature Selection for Simple Color Histogram Filter based on Retinal Fundus Images for Diabetic Retinopathy Recognition
Published in IETE Journal of Research, 2023
T. Vijayan, M. Sangeetha, A. Kumaravel, B. Karthik
Diabetic Retinopathy is best diagnosed with a comprehensive dilated eye exam and fundus photography. Fluorescein angiography and Optical Coherence Tomography (OCT) are also performed for better decision making especially for Diabetic maculopathy and proliferative retinopathy. Decisions taken by a comprehensive dilated eye exam and fundus photography are subjective and the sensitivity and specificity are varying from one another. The practice of e-maintenance of health data and the evolution of artificial intelligence and deep learning algorithms [12–15] are supporting physicians to improve the decision making in this context. In recent years, the research community has developed different techniques and applied several deep learning frameworks to address this problem. The input images for learning models are fundus images with labels indicating the presence or the level of severity of Diabetic Retinopathy identified by professional graders. We apply supervised learning models like decision trees and k-nearest neighborhood for this purpose.
An intelligent approach for detection and grading of diabetic retinopathy and diabetic macular edema using retinal images
Published in Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization, 2023
Pranoti Nage, Sanjay Shitole, Manesh Kokare
Diabetic retinopathy (DR) is caused due to a disorder named diabetes mellitus which can damage the retina and even lead to loss of vision. The DR has several stages of severity such as mild, moderate and severe. The severe stage of DR is termed proliferative diabetic retinopathy (PDR) in which the formation of new vessels in the retina is observed (Sugeno et al. 2021). However, the early detection of DR and proper diagnosis will reverse or reduce the growth of the effects caused by the disease (Acharya and Kumar 2021). Diabetic macular oedema (DME) is a condition in which the lesions caused by DR are observed in the middle portion of the retina called the macula. The DME is considered a severe condition as the damages caused by it is irreversible (Lxcamey and Palma 2021). The detection of these diseases is carried out by identifying features such as micro-aneurysms, hard exudates and haemorrhages, as illustrated in Figure 1. The micro-aneurysms refer to the red spots in the retina’s blood vessels with sharp margins, retina’s blood vessels with sharp margins formed in the early stages of the disease (Washburn et al. 2020). The exudates are caused due to abnormalities in the blood vessels which are formed as yellowish-white spots in the outer layer of retina (Wang et al. 2020). The haemorrhages also occur like micro-aneurysms but have irregular margins caused due to the leakage of capillaries which is a delicate blood vessels (Kanimozhi et al. 2021). The blockage of arteries also contributes to a condition named cotton wool spot, which occurs as a white region in the retinal nerve (Chaudhary et al. 2021).
Diagnosis of Covid-19 using Chest X-ray Images using Ensemble Model
Published in IETE Journal of Research, 2023
K.V. Uma, C. Sakthi Birundha, S. Subasri, V. A. Harini
A method was proposed based on the attention module and the pre-trained Deep learning model (VGG-16). The VGG-16 model was chosen for two reasons. First, it uses its smaller kernel size to extract features at the low level, making it suitable for CXR images with fewer layers than its VGG-19 counterpart. Second, it is better at extracting features for COVID-19 CXR image classification. The fine-tuning strategy is employed here. They employ the Image Net pre-trained weight for fine-tuning with the VGG-16 model. It aids in overcoming the issue of over-fitting [11]. The classification of COVID-19 cough sounds was done with the help of Machine learning algorithms for the Coswara dataset, which consists of 480 healthy and 160 infected patients. Data is collected through the non-contact-based screening test performed on the individuals within their residence boundaries. Different classification algorithms are applied to differentiate the COVID-19 coughs and healthy coughs. The SVM algorithm provided better classification results than other algorithms [13]. A GIL-CNN model was proposed to categorize CT scan images of COVID-affected patients from healthy persons. The model uses three kinds of features, global, local, and intermediate, for the efficient feature representation of CT scan images. By feeding GIL-CNN features as input for Support Vector Machine, normal and Covid images are classified. The proposed model provides better performance and can be used to identify Covid-19 patients in an automated manner [14]. A new framework that involves a four-step feature selection technique and an ensemble deep neural network was proposed to detect one of the eye diseases called diabetic retinopathy automatically. Diabetic retinopathy, if untreated, may lead to permanent vision loss. The features from retinal fundus images are extracted using Inception V3, RestNet101, and VGG19. The framework comprising four efficient feature selection techniques reduces the image's feature space. Then the feature space is classified using Support Vector [15].