A Wearable System with Real-Time Diagnosis of Arrhythmia based on Fuzzy Algorithm
Most arrhythmias are hard to detect in the early stage without regular health diagnostics, and often this leads to
patients discovering arrhythmia after they had already reach a lethal stage. If one could diagnose and analyze the data in
advance to start the treatment of arrhythmia, there would be less cases that fail to benefit from the golden treatment period.
This paper develops an automatic recognizing arrhythmia electrocardiogram (ECG) signal measurement and monitoring
system, using the light and wearable device to extract single-lead ECG signal and transmit through Bluetooth to the system
receiver. The system uses Dynamic Threshold Crest Detection (DTCD) method to calculate ECG time sequence feature values
then convert and analyze to acquire the average heart rate, QRS complex duration, and heart rate variability (HRV) for fuzzy
system input parameter. The data are classified by fuzzy inference and degree of membership analysis. The inference of this
system is not required to build complicated mathematical models and training samples, thus it possesses a simple and rapid
advantage. According to the experiment results, this system can be effectively used to diagnose 5 types of cardiac arrhythmia
and give feedback to users immediately. Data stored can be applied as the diagnostic assessment indicators for physicians in
order to achieve the goal of disease prevention.