A Soft Computing-Based Measurement System for Medical Applications in Diagnosis of Cardiac Arrhythmias by ECG Signals Analysis

In this paper a telemedicine application in support of heart patients homecare is proposed. Diagnosis of cardiac arrhythmias typically requires heart monitoring for 24 or 48 consecutive hours, so it causes long waiting lists in hospital. In this sight the authors have projected an user-friendly virtual assistant able to acquire and analyze patient ECG signals by soft computing procedures. The system is based on a PDA equipped with a DAQ card acquiring, to a fixed sampling frequency, the heart electrical impulses by means of an ECG sensor. Then an embedded algorithm allows to depict the graph of the electrocardiogram on the PDA display. By build-in models the ECG waveform is analyzed in order to diagnose possible arrhythmias occurrences or the happening of a heart attack. In fact, the developed computational intelligence application enables the system to perform a patterns recognition taking into account information on the measurement uncertainty. In this way, according to some patient parameters like age, sex and physical constitution, a set of warning lights on display provides information on the current heart status of the patient. Consequently the heart report can be sent by a GPRS modem to a confidential Web page containing patient personal data. The M2M application allows information to be made available to expert medical staff of hospital or clinic for a further remote analysis. In presence of a potential emergency, an online doctor can decide the typology of intervention for the patient assistance