Diabetic retinopathy is one of the most frequent complications diabetes, which leads to severe consequences, including rapid and irreversible vision loss. The laser coagulation procedure treat diabetic consists in applying a series microburns fundus deal with macular edema. existing hardware/software packages are primarily based on use predetermined pattern for coagulate arrangement. However, due complex form edema vascular system, this approach an uneven To solve problem, we propose new...

The paper proposes a technology for effective feature selection to localize individual characteristics of anatomical and pathological structures in the human eye fundus. Such an approach allows intellectual analysis features be conducted using color subspaces regions interest identified. This problem is relevant because this way efficiency laser coagulation surgery can improved. based on texture certain image patterns. initial textural attributes are derived from different statistical...

For diabetic retinopathy treatment, laser coagulation is used in modern practice. During the surgery process, parameters of exposure are selected manually by a doctor, which requires doctor to have sufficient experience and knowledge achieve therapeutic effect. On basis mathematical modeling it possible estimate crucial without performing an operation. However, retina has rather complex structure, when even low-cost numerical methods for modeling, takes long time obtain result. In this...

A personalized medical approach can make diabetic retinopathy treatment more effective. To select effective methods of treatment, deep analysis and diagnostic data a patient’s fundus are required. For this purpose, flat optical coherence tomography images used to restore the three-dimensional structure fundus. Heat propagation through is simulated via numerical methods. The article proposes algorithms for smooth segmentation retina 3D model reconstruction mathematical modeling laser exposure...

The paper proposes a method for selection the region of diabetic macular edema in fundus images using OCT data. relevance work is due to need create support systems laser coagulation increase its effectiveness. proposed algorithm based on set image segmentation methods, as well searching specific points and compiling their descriptors. Canny used find boundary between vitreous body retina images. method, Kruskal constructing minimum spanning tree weighted connected undirected graph, select...

In the analysis of a crystalline substance, problem crystal lattice parameter identification is great interest. However, existing methods for solving this problem, such as Bravais cell parameters estimation method and Wigner-Seitz volume method, do not provide required level accuracy. Aiming to address low accuracy, paper proposes an algorithm based on gradient steepest descent method. The study feasibility structure carried out using large set distorted lattices. results obtained show...

Abstract In this paper the study of fundus image segmentation using convolutional neural networks is carried out. A network architecture was made to classify four classes images, which are up thick and thin blood vessels, healthy areas, exudate areas. The CNN constructed empirically so as required accuracy no less than 96 % ensured. error calculated on exudates class, key for laser coagulation surgery. we utilized HSL color model because it renders characteristics eye vessels most...

The article deals with a problem of three-dimensional crystal lattice reconstruction, which is an important stage in the X-ray structural analysis. accuracy parametric and identification crystals directly depends on quality reconstruction. proposed algorithm reconstruction based minimizing distances from each node to line projected onto specified plane. Three sets two-dimensional coordinates, obtained three projections, are used as input data. We performed analytical calculation error,...

Aim: to compare the uniformity and adequacy of placement laser spots after mono-impulse pattern photocoagulation for diabetic macular edema (DME). Patients Methods: fundus photographs 83 patients (121 eyes) taken right retinal DME were analyzed. Group 1 included images 63 eyes group 2 58 photocoagulation. Laser burns varying intensity based on LʹEsperance scale (including grade 0 that not seen photos) calculated. Grade considered optimal. The number non-optimal placed hemorrhages, blood...

The eye's fundus is a complex object with high density of filling vessels. procedure treatment requires lot experience and skill from the physician. basic tool in laser coagulation, when series metered microscopic thermal wounds – coagulates are applied edema on retina. Currently available software packages focused first use specified template (pattern) for coagulate location. Effects pattern application caused by irregular location reason variability vasculature forms. paper presents...

To increase the efficiency of laser coagulation surgery problem most accurate segmentation fundus images is especially relevant. Fundus image carried out with high accuracy using effective features and minimum number parameters for a single fragment. This paper describes modified technique smart textural feature selection to extract retinal regions interest preprocessing algorithms. Preprocessing algorithms significantly influence selected which provide error object recognition. In addition...

When laser coagulation of eye retina is carried out, the beam directed to target retinal areas selected by an ophthalmologist. The exposure light produces a photocoagulate. using coagulation, main problem selecting both that would cover all pathological zones and parameters prevent damage. Any patient has individual fundus structure. structure anatomical elements must be taken into account achieve desired therapeutic effect. formation coagulates in hemorrhage-affected results vascular layer...

Diabetic retinopathy is among the most severe complications of diabetes, often leading to rapid and irreversible vision loss. The laser coagulation procedure, which consists applying microburns fundus, has proven be an effective method for treating diabetic retinopathy. Unfortunately, modern research does not pay enough attention study arrangement in edema area—One key factors affecting quality therapy. aim this was propose a computational decision-making support system retina...

Abstract This paper proposes a technology for lung image segmentation using textural features. To analyze the informativeness of features, K-Means method was used. The result can be used to describe characteristics patient: age, gender, physique, etc. study conducted on large set fragmented images. X-ray clustering errors 12×12, 24×24, and 36×36 fragmentation windows were presented. Pre-processing images also estimate quality. showed that provided key objects selecting error at little more...

Abstract We propose a technique for localization of diabetic macular edema areas via graph-based segmentation OCT retinal images. The relevance the research is associated with high incidence rate severe eye conditions due to among world population. relies upon highly efficient image segmentation. Using set specially selected parameters, accuracy area enhanced. Optimal parameters found in course have enabled error 2% be achieved.

Abstract This paper proposes the technology for large biomedical data analysis based on CUDA computation. The was used to analyze a set of fundus images diabetic retinopathy automatic diagnostics. A high-performance algorithm that calculates effective textural characteristics medical image has been developed. During diagnostics, following classes were distinguished: thin vessels, thick exudates and healthy area. study mentioned efficiency conducted with 500x500-1000x1000 pixels using square...

In this work, we proposed a new approach to analyzing eye fundus images that relies upon the use of convolutional neural network (CNN). The CNN architecture was constructed, followed by learning on balanced dataset composed four classes images, thick and thin blood vessels, healthy areas, exudate areas. conducted 12x12 because an experimental study showed them be optimal for purpose. test error no higher than 4% all sizes samples. Segmentation performed using CNN. Considering exudates are...

Abstract The study of substances with a crystal structure is complex multi-step process. key step in the crystalline substance analysis unit cell parameter estimation. estimation lattice parameters particular problem that involves search model’s according to information which can be extracted from substance. In these recent times, most accurate about obtained electron microscope whose linear resolution high enough observe atomic this case means reconstruction three-dimensional 2-dimentional...