A System For Real-time ECG Detection And Analysis Based On AT91SAM9X35

With the accelerated pace of society, cardiovascular disease has become the first killer of humans. In order to effectively monitor patients with cardiovascular disease, in the 1960s, Doctor Holter successfully developed Holter System. With the development of modern communication technology, Holter system evolved into ECG monitoring system based on Internet and ECG monitoring system base on wireless communications, and now ECG monitoring system based on CDMA and Internet and ECG monitoring system based on GPRS and Internet.As ECG monitoring system’s terminal, ECG monitors mainly complete functions including data acquisition, storage, transmission and analysis. Constrainted by monitor equipments, algorithms implemented under the embedded environment not only need good effect but also features of simple and real-time. Currently ECG algorithms implemented under the embedded environment mainly focus on the detection of QRS wave and heart rate calculation based on the detection of QRS wave. P wave, T wave and ST segment detection are relatively difficult, and the accuracy rate is not high.To this end, we built an embedded linux operating system base on board chip AT91SAM9x35, and then built a QT platform in the embedded linux operating system, and then developed ECG detection and analysis system using QT Creator in PC. After the development of ECG detection and analysis system, we used qmake to make executable files and transfer these executable files under the network file system rootfs, and running in the embedded environment.Our ECG detection and analysis system use a number of internationally recognized ECG database, including the MIT-BIH database, AHA database, and ST-T database. WFDB library is an good tool to use these online ECG databases, we integrated WFDB library functions into our system to analyze algorithms in our system and finally give the results. The results show that the real-time ECG detection and analysis algorithm used in our system has features of high accuracy and good real-time, and this algorithm is suitable for real-time detection of ECG under embedded environment.The system consists of four threads, namely acquisition thread, main thread, analysis thread and display thread. Acquisition thread is responsible for data collection. When using MIT-BIH database, we use an analog acquisition thread to read ECG data from files, and then transmit the ECG data to the main thread. After received ECG data, the main thread process data to calculate the ECG parameters, and then transmit the ECG data and ECG parameters to the display thread. The analysis thread is triggered by the button on the interface, after user clicks the button, the analysis thread end the acquisition thread and the main thread, and then analyze the results of the main thread, and then transmit the analysis results to the display thread. Communication between threads is accomplished by the slots mechanism of QT.