A. S. Krasichkov

The practice of modern ECG monitoring devices application in alarming an ischemia patient has demonstrated their impaired capability in clinical use due to the high false alarm rates that appear a significant psychological factor. Therefore it is essential to considerably improve the accuracy of the alarm development in the device that performs 10000 evaluations every 24h. The conventional methods to synthesize monitoring devices with implemented ECG online processing algorithms are oriented to a general user. This is in contrast to the current tendency in cardiology that appears the individual evaluation and management of every patient. Based on the analysis of modern clinical requirements, current state of the available equipment and biological signal analysis methods an approach to synthesize individual ECG processing algorithms under the assumption that the diagnosis is clinically approved and a holter moniroting record and functional load test records obtained under clinical conditions are available. The duration of the available clinical records should be sufficient to estimate the ECG parameters with required accuracy. Next, based on the suggested approach an individual algorithm is synthesized and can be used in the monitoring device with alarm features. The decision set of the algorithm consists of two possible values: alarm and non-alarm, the parametrization is based on the individual myocardium ischemisation level, and ECG realizations form the decision statistics.