Search academic texts

Information × Registration Number 0223U001881, 0122U200266 , R & D reports Title To develop an intelligent system of automated evaluation of tumor tissue characteristics by microphoto. Step 2. To develop software for manual marking of breast microphotographs in order to determine the characteristic areas for the primary selection of informative and non-informative fragments of medical images of breast cancer integrated with BDF breast cancer. To determine a set of source data (attributes) of informative fragments of microphotographs of breast cancer for presentation to the system of diagnostic models, to develop and implement algorithms for their calculation. popup.stage_title Head Tulchynskyi Vadym Н., д.ф.-м.н. Registration Date 04-02-2023 Organization V.M. Glushkov Institute of Cybernetics of the National Academy of Sciences of Ukraine popup.description2 ABSTRACT Scientific Research Report: 32 pg., 11 pic., 0 tabl., 24 references BREAST CANCER, HISTOLOGICAL MEDICAL IMAGES, MICROPHOTO DATABASE, BINARY SPARSE DISTRIBUTED REPRESENTATIONS OF SEMANTICS, ASSOCIATIVE MEMORY, NEURAL NETWORKS The project goal is to automate the process of breast cancer (BC) diagnosing by developing an intelligent analyzer of morphological features of breast cancer cells and tissues based on histological medical images (microphotographs) for use in scientific research. The project tasks: to develop software tools for manual segmenting BC microphotographs in order to determine the characteristic zones for the primary selection of fragments of BC medical images integrated with the database of breast cancer microphotographs (DBMP); to select the initial set of attributes for informative fragments of the BC microphotographs as the input data for the future system of diagnostic models; to develop and implement algorithms for computing the attributes. The main obtained results: 1) the developed manual segmentation tools, that turned out to be unnecessary, because according to additional research it is only possible to classify a posteriori entire microphotographs; 2) two collections of microphotographs of totally 1789 pictures (1076 and 713, respectively) from two different microscopes accumulated and classified according to the degree of BS aggressiveness (done by the partners from IEPOR NASU); 3) the image collections were relocated to a foreign hosting to bypass the periodic power outages due to Russian attacks, and to fit the hosting limitations, a new web service and web application have been developed in PHP for browsing the collections and providing remote access to them; 4) a set of invariant attributes (spectrum of lineament angle distribution, histograms of lineament lengths and thicknesses) and 2 Python programs developed for the attribute computing (an interactive one in Jupiter Notebook and a batch one to compute a sample set ) Product Description popup.authors Bilodid Oksana O. Bandura Olecsandr Y. Vdovichenko Ruslan O Golovinskiy Andriy L. Lavrenuk Sergiy I. Nazarenko Yeмпут М. Prazdnikova Mariya O. Roganov Viacheslav Yu Sidoruk Volodumir А. Tulchinskiy Vadym H. Tulchynskye Petro H. Halimendik Valeriy. V. popup.nrat_date 2023-02-04 Close