Advanced Information Fusion Approach to Heart Vessel Modelling and Recognition

Methods of multi-criteria information-based reasoning represent a major topic in the knowledge engineering and data aggregation techniques currently developed. Such techniques can be of great significance and find a number of applications in image recognition and determining the semantics of visual patterns. One of the most important fields in which such methods could be applied, is the analysis of medical images showing the complicated morphology of selected structures or organs. This study presents a proposal for the semantic analysis of coronary vascularisation, carried out using 3D reconstructions obtained in spiral CT examinations. Such an analysis would be comprehensive and consist in aggregating diagnostically significant information about the morphology of individual vessels, their topological relationships and functional condition.