Design And Implementation Of An Intelligent Domestic Multi-function ECG Monitor Based On ARM Platform

With the ability of recording long time ECG data, ECG monitor has become important equipment in hospital. In recent years the home medical device market over the world is booming. The problems of rapidly ageing populations, stress of city life, and disease prevention have driven the market internally. The rapid development of medical detection technology and computer technology also put forward the growing trend.Family ECG monitor is derived from the hospital monitor. Its manufacturers mainly focus on such improvements like wireless network, portability, simplifying functions, and optimizing operation. However some other demands of family users are still not satisfied. ECG intelligent detection function is the most wanted. So this paper design and develop an intelligent system for family ECG monitor.ECG detection algorithm is core of this system. A method of machine learning is used in this paper. ECG data is quite different from general machine learning data, for it is shown in the form of wave. So, the first step is to extract valuable data for learning. Wavelet coefficients from quadratic spline wavelet transformation with compact support and one vanishing moment are used to detect R wave, the most notable wave in electrocardiogram. According to R wave’s position, single heart periods are extracted from long time electrocardiogram. Then A3wavelet coefficients are used as training vectors for machine learning. This paper selects6kinds of ECG waveform from the MIT-BIH database, and then training a decision tree model, which perform high accuracy rate.Second part of this paper introduces the realization of software system. The system is based on ARM platform which becomes the most popular platform in electronic market, Linux and Qt GUI. The system follows the MVC design pattern to simplify its architecture and to make maintenance convenient. This paper expatiate the realization of5main functions, including serial transmission, graphic interface, alarm, parameter setting and intelligent ECG detection.