Knowledge-based approach to the management of serious arrhythmia in the CCU.

An expert system (SETA) for the management of patients in the CCU environment has been developed. SETA suggests therapeutic actions for the treatment of serious arrhythmias that complicate the pathophysiological state of patients recovering from acute and suspected myocardial infarction. The prototype begins by reasoning from the arrhythmia, diagnosed from the ECG signal, and progresses through its inference process by considering ECG changes (i.e. heart rate, QRS width), patient clinical data, patient history and therapeutic drug data, to reach the most appropriate actions for each particular patient. The system was implemented on production rules using M1 as a development tool. SETA uses a multiknowledge base (KB) architecture, one for each particular arrhythmia and relevant complication, and a decision board that controls the firing of the KBs and keeps track of patient status through time. The system takes into consideration aspects that are very important for the human expert, e.g. sequence of arrhythmia appearance, drug contraindications and priority in the case of simultaneous arrhythmia. The development of this system has given insight into the management of critical CCU patients, that should influence the specification and design of intelligent instruments for this clinical environment.