Makhno T. Automated system for processing ultrasound images of the carotid artery based on evolutionary algorithms

Object: methods of ultrasound medical image processing. Methods: genetic algorithms, genetic programming, methods of digital image processing, systems analysis. Aim: to increase processing efficiency of ultrasound images of human carotid arteries. Theoretical and practical results: solution of an actual scientific and practical task of designing information technology of ultrasound images processing based on evolutionary algorithms that improves the efficiency of ultrasonic image segmentation. Scientific topicality: a new genetic programming-based method of synthesis of ultrasound image segmentation schemes that increases the rate of synthesis of accurate segmentation schemes due to proposed modifications of crossover and mutation operators was developed; a method of reducing the set of image processing algorithms being used was offered that allows to improve the solution of medical ultrasound images segmentation task by way of accelerating process of synthesis of accurate segmentation schemes; the technology of ultrasonic imaging has been further developed through the automation of image segmentation methods selection that depends on a set of automatically computed values for texture parameters chosen for preliminary appraisal of ultrasound images. Degree of implementation: The results of thesis research were implemented in The Institute Of Urgent And Reconstructive Surgery of National Academy of Medical Sciences of Ukraine. Area of use: Medicine.