Embedded Real-time ECG Diagnosis Algorithm

This paper discusses a set of ECG diagnostic algorithms used in the embedded monitor. From reading the ECG data, pre-processing to a variety of waveform parameters calculation and analysis, the real-time ECG monitoring under embedded environment is achieved. The database used here is MIT-BIH Arrhythmia Database, the programming language is C, and the algorithm is realized in the QT programming environment under Linux, with which the result can be displayed on the screen, while effects of the algorithm are counted in the MATLAB environment.Firstly, with the introduction of Current Status of ECG monitor and developing trends of real-time ECG diagnosis algorithm, here comes the importance and necessity of ECG algorithm applied in embedded monitor. The next part introduces the overall framework of ECG monitor and the application background of ECG algorithm, which is in the application layer module of the whole system, and also ways of showing results of the algorithm, all of which supply foundation of the specific design of the algorithm below.Then, from the introduction of details of the database used in this paper, including data sources, data reading and data authorities, the paper discussed the whole set of ECG diagnostic algorithms, including data pre-processing and follow-up analysis. For Pre-filter of ECG, here analyses different design methods of the low-pass filter, high pass filter and filter for 50Hz interference and gives the summary. The following is wave testing and analysis, including feature waves extraction, marking and calculation of various wave parameters.Finally, display effects of the algorithm applied in embedded monitor are given, while effect of the set of algorithms were analyzed and summarized. Following parts give summary of this article, the advantages and disadvantages of this set of algorithms and the future developments.The ECG algorithm achieves a good effect for real-time monitoring. R wave detection accuracy rate here is 99.422% according to the statistical test results of part data of the MIT-BIH Arrhythmia Database, while the waveform detection and parameters calculation also reaches good levels. The algorithm is simple and effective, which is suitable for use on embedded monitor, however, it still cannot diagnose disease in deep details by in-depth analysis and diagnosis and can be further improved.