EcgRuleML: A rule-based Markup Language for describing diagnostic ECG criteria

From the clinician´s perspective, computerized interpretation of the electrocardiogram (ECG) is carried out in a `black box´. That is, the rules used for interpretation are not easily accessible to the clinician. In this study we propose the ECG Rule Markup Language (ecgRuleML) as a way to externalize decision rules used to interpret the ECG. EcgRuleML utilizes the eXtensible Markup Language (XML) to provide a framework for articulating quantitative rules for measuring intervals, segments, widths, peaks, heart rate and the cardiac axis. Abstract features of the ECG such as slurred S waves cannot be easily represented numerically and are therefore articulated using codes. To test the ecgRuleML framework, rules have been defined to assess ST Elevation Myocardial Infarction (STEMI) in a Lux-192 Body Surface Potential Map (BSPM). An algorithm has been integrated into a BSPM viewer where the rules have been parsed from an ecgRuleML document and executed in 63ms (mean from 10 trials) on a PC (3GHz CPU, 3GB RAM).