Medical and biological images have different physical natures. A person-operator can perceive them in different ways. Therefore, an important task is the adaptation of the image processing process to a specific user, that is, to a narrow task that is solved by the consumer of information. It is often not enough to present an object to an observer with the help of an ideal display system, since the information required for image analysis in order to search and identify objects, the determination of various quantitative characteristics can be manifested only as a result of digital processing. Based on the above, the paper proposes an automated method for extracting the contours of ultrasound images using the example of studying the state of the hip joint. This method makes it possible to construct conventional lines on the image of the hip joint in an automated mode with minimal expenditure of resources and subsequently to calculate the values of certain angular indicators, and using known methods of studying the condition of the hip joint, to estimate its degree of dysplasia. By automating the process, the reliability of diagnostics can be increased.
|