Search academic texts

Information × Registration Number 0226U001613, (0121U108937) , R & D reports Title Development of methods, algorithms and software for synthesis and classification of biomedical images popup.stage_title Розробка методів, алгоритмів та програмних засобів синтезу та класифікації біомедичних зображень. Head Berezkyi Oleh M., д.т.н. Registration Date 28-01-2026 Organization Western Ukrainian National University popup.description1 The aim of the work is to develop methods, algorithms and software for the synthesis and classification of biomedical images. popup.description2 The object of the study is the processes of automated analysis, synthesis and classification of biomedical light microscopy images (histological, cytological and immunohistochemical) using computer vision and artificial intelligence methods. The subject of the study is methods, algorithms and metrics for assessing the quality of biomedical images at low, medium and high levels of computer vision, as well as algorithmic approaches to the synthesis and classification of biomedical images based on deep neural networks. The goal is to develop and systematize methods for assessing the quality, synthesis and classification of biomedical images taking into account their structural, statistical and semantic properties in order to increase the reliability and reproducibility of automated analysis systems in biomedical and clinical applications. The goal set has led to a number of research tasks: to analyze the physical and morphological features of the formation of histological, cytological and immunohistochemical images; to investigate the structural and statistical properties of biomedical images and the sources of their variability; systematize the levels of analysis in computer vision for biomedical data; consider and classify low-, medium-, and high-level metrics for assessing the quality of biomedical images; analyze the application of the Fréchet metric in biomedical image synthesis problems; investigate the Hausdorff metric and its modifications for assessing the quality of segmentation; consider approaches to the use of fuzzy metrics and fuzzy similarity; analyze modern methods and algorithms for biomedical image synthesis, in particular based on deep neural networks, generative adversarial and diffusion models; generalize methods and algorithms for classifying biomedical images using deep neural networks and convolutional architectures. Product Description popup.authors Yuriy M. Batko Hryhorii M. Melnyk Petro B. Liashchynskyy Pavlo B. Liashchynskyi Oleh Y. Pitsun Mykola O. Berezkyi popup.nrat_date 2026-01-28 Close