Intelligent Processing Inapplication Research Of EGC Detection

The research of electrocardiogram（ECG for short）is a hotspot in intelligentprocessing at present,because cardiac disease is a serious threat to human health and isa killer to human health,the analysis of ECG is an important approach for cardiacdisease.it will improve the development of medical cause,the level of people’s healthand have an important significance to diagnose heart disease,if achieving intelligentprocessing of ECG signal.Presently there are lots of problems including thedenoising,the feature extraction and the classification of ECG signal in intelligentprocessing of ECG.Based on previous research,this paper focuses on the research ofintelligent processing on the above three aspects.the main task reaearch andinnovation are concluded as follows:The research of the denoising of ECG signal: the denoising is not only thebasis,but also the pivotal step.The main noise include frequency disturbance,baselinewandering,muscle artifact.According to the noise’s feature,designing a method ofdenoising based on morphology and wavelet transform.Experiments show that thismethod can denoise noise existing in ECG signal.The research of the feature extraction of ECG signal: the feature extraction is im-portant and difficult for diagnosingArrhythmia rhythm.Because of ECG signal is co-mplicated,the detection precision influences the result of classfication,and QRS dete-ction is crucial in all the feature extraction and is the base of other feature extraction,Designing a method based on mophological self-adaptive to detect QRS in this pape-r.Experiments show that the accuracy is high.The research of the classification of ECG signal: Anovel approach is put forwardfor classifying Arrhythmia rhythm signals based on template matching and SupportVector Machine(SVM).The article mainly aims at classification of four kinds of usualArrhythmia rhythm heart beat and normal heart beat.The template width of themethod constantly upate follow with the signal changes and the signals of the templatechange,increasing morphological difference signals,the classficiation results are well.The algorithms are designed on Matlab platform,and evaluated using a part ofMIT-BIH Arrhythmia rhythm database. The intelligent processing of ECG signal is avery huge task, needing research and improvement continously.