Search academic texts

Information × Registration Number 0217U000300, 0116U002500 , R & D reports Title Hybrid intelligent information technology diagnosing precancerous breast cancer based on image analysis popup.stage_title Head Berezkuy Oleg Nukolaevich, Доктор технічних наук Registration Date 26-01-2017 Organization Ternopol National Economic University popup.description2 Object of study - hybrid intelligent information technology. Purpose - to develop hybrid intellectual information technology and intellectual system of diagnosing precancerous breast cancer based on analysis of histological and cytological images. Methods: methods of the theory of discrete groups of crystallographic methods topology, methods of linear algebra and analytic geometry, methods of computer vision, methods of computer modeling. Results: methods and algorithms for pre-processing and image segmentation, image processing algorithms on GPUs, fuzzy knowledge representation models, algorithms of information protection in automated microscopy systems, components intellectual system of diagnosing precancerous lesions. Key Features of software: development language Java, PHP, development environment Intellij IDEA, algorithms, computer vision library OpenCV, operating system, Microsoft Windows and Linux, the size distribution of 20 MB. Composition hardware subsystem input image, digital camcorder TCA-10, Bresser. Key Features TCA-10 camera, the signal / noise ratio of 40 dB, 60 dB dynamic range, image resolution is 3264x2448. Key Features microscope Bresser: Working biological microscope transmitted light, increasing the range of 40 degree 1000. Implementation: report on the research work. Areas of implementation: the work can be used in clinical practice for research in morphology and in the classroom. Field of application - medicine, histology, cytology. Cost-effectiveness: the use of systems designed to increase productivity cytologists, patohistoloha and reduce complexity morphometric analysis. The significance of the work and findings, methods and algorithms convert images into a topological space, methods and algorithms for evaluation of image quality in the metric fresh, Hausdorff, Gromov-Hausdorff method and algorithms suspended chords to match areas of images, the method and adaptive algorithms median and Gaussian filtering method Count algorithm development process images on a GPU with multiport memory. The estimated directions of research: the study of new types of tumors, development of knowledge bases for staging aided diagnosis, developing intelligent systems for automated microscopy and cytologists patohistoloha, developing technology to protect computer systems, components development of telemedicine. Product Description popup.authors Ігнатєв Ігор Васильович Батько Юрій Мирославович Березька Катерина Миколаївна Боднар Андрій Русланович Вербовий Сергій Олегович Гумовська Ірина Миколаївна Дацко Тамара Вікторівна Дубчак Леся Орестівна Зубко Віталіна Сергіївна Лящинський Павло Борисович Лящинський Петро Борисович Мартинчук Тетяна Олегівна Мельник Григорій Миколайович Піцун Олег Йосипович Савка Надія Ярославівна Цмоць Іван Григорович popup.nrat_date 2020-04-02 Close