A computerized method for the application of coronary artery disease risk classification models to a large patient population

The authors developed a flexible and powerful coronary anatomy classification system which can be applied to the development and evaluation of qualitative and quantitative methods for assessing myocardium in jeopardy on the basis of the severity and location of coronary artery lesions, relative size of arteries, and contributions from collateral blood flow. The system employs a graphical user interface for the interactive generation of a coronary tree diagram depicting a patient´s coronary anatomy. An anatomical relational database for over 5000 patients has been collected. Various classification schemes can be rapidly applied to this data set and iteratively modified to arrive at a model which correlates best with patient prognosis. As an example of this system´s ability to easily implement complex coronary artery disease risk classification models, a risk region analysis model based on size and distribution of terminating vessels was implemented. This method offers a significant improvement over its predecessors in that it takes into account changes in coronary size and distribution among different patients