Linguistic Approach to Modeling of Coronary Arteries in Semantic Techniques of Image Retrieval

In recent years, medical images have become a constant part of our life. It is hard to imagine making a medical diagnosis without basic medical imaging apparatuses (CT, MRI, USG, etc.). In order to fully utilize its potential, the image data must be stored in the proper way, particularly so that it is easily accessible. The latter is particularly important in a medical diagnostics - it is desired that a physician should have quick access to the comparative image data. In this context the authors present the possibility of using linguistic approach to modeling of coronary arteries in semantic techniques of image retrieval. The proposed methodology takes into account, among others, the basic topologies of arteries (balanced artery distribution, right dominant or left dominant) as well as individual sections or ramifications of these arteries. For the purpose of analysis and indexing these visualizations appropriate graph and sequential grammars are proposed.