MODELS AND INFORMATION TECHNOLOGY OF PROCESSING ULTRASOUND IMAGES

Abstract

Three-dimensional ultrasound images offer much more information compared to conventional two-dimensional images, allowing doctors to “rotate” and “flip” the image to explore it from all possible angles. This level of detail can be critical for detecting pathologies or planning surgical interventions. However, effectively working with such images requires specialized software, models, and information technologies capable of processing large volumes of data and reproducing them as intuitive 3D models. Developing this kind of software requires consideration of numerous technical and medical aspects. The purpose is to analyze the models and information technologies for processing ultrasound images, which are of great importance in medical diagnostics. The article considers and compares various methods and models that can provide the best results for image processing. The methodology involves the following scientific research methods: analysis, synthesis, comparison, and generalization to consider the main aspects of the studied problem and define the theoretical foundations of the research. The article reviews the main approaches to modeling the processes of ultrasound image processing, including methods for improving image quality, extracting contours, and structural elements. Algorithms and software tools that ensure effective processing and analysis of ultrasound data are described. The scientific novelty of the results obtained in the work lies in the formulation and use of the best methods, models, and information technologies for processing ultrasound images, which are of great significance in medical diagnostics. Attention is given to noise filtering methods, image contrast enhancement, and automatic pathology recognition. The research results demonstrating the effectiveness of the proposed approaches in clinical practice are presented. Conclusions. The methods reviewed simplify the use of modern algorithms for automated analysis of ultrasound images. The review of available tools shows that 3D Slicer offers researchers convenient graphical and software interfaces that facilitate the implementation and application of the latest machine learning algorithms in ultrasound image processing. The article also discusses the prospects for further development of information technologies in ultrasound diagnostics.