[Ergometric, hemorheologic and myocardial perfusion studies and their evaluation by pattern recognition method in...

A large number of noninvasive and invasive methods are used in establishing the diagnosis of ischemic heart disease. Coronary angiography is an expensive method and a restricted capacity exists in Hungary. The aim was to elaborate an optimal combination of different noninvasive methods which can select the real positive cases for coronary angiography with maximal efficiency. Besides routine examinations, dobutamine pharmacological test was carried out and hemorheological parameters were also determined. After evaluating the results, patients were divided into three groups: absence of IHD, probable IHD and IHD. The clustering of patients into these groups was made by using their multivariable classifier algorithm (PRIMA) and 27 different parameters of 44 patients were taken into account. The groups were well separable as the class-distances indicated. The maximal ST segment depression during exercise, pharmacological test and hemorheological parameters were found to be the most powerful discriminating factors. Though the validation of the results is still in progress, based on their previous data they feel that the examination of hemorheological parameters and the application of pattern recognition method can be useful in the diagnosis of ischemic heart disease.