A Technology In The Detection Of Transient Nonlinear Alternans Wave

In recent years many studies have shown, T-wave alternans(TWA) is that theshape of T wave appear alternately which is one high T wave, one low T wave inelectrocardiogram. It has already been used as the effects of malignant arrhythmia andsudden cardiac death independent predictor. This research can prevent the suddendeath, and this research has played a very important role to protect people’s health.However, it is very difficult to observe whether there is any alternate T bandphenomenon with the naked eye in electrocardiogram, because most of the T wavealternation phenomena are slightly level, So we must use the modern signal andinformation processing technology to analysis detection operation whether there is theT wave alternating phenomenon.There are many ways to detect the phenomenon of T wave alternans, based onthe degree of easy operation and excellent real-time performance point of view, in thispaper we select the correlation analysis of time domain analysis method to analyze thedetection of T wave alternate phenomenon, This paper mainly introduces thefollowing work:(1)Firstly, this paper introduces the most basic parts, the composition of theelectrocardiosignal, the requency range of each band, make sure the definition of thephenomena of T wave alternates clearly, and the happening mechanism of T wavealternans.(2) T wave alternate of accurate analysis detection, for the pretreatment of theoriginal signal, the denoising is indispensable part, this paper described several kindsof main noise type, and a more effective and frequently-used denoising method, inthis article,based on considering of the real-time and calculation, we choose the useof simple integer coefficient method, FIR filter and the method of combining thewavelet transform to get rid of the several kinds of noise in electrocardiosignal mainlyexisting: power frequency interference, emg interference and baseline drift.(3) Feature point detection and calibration, because the R wave is the part ofstrongest observational, and highest frequency, accurate measurement and calibrationof R peak is relatively easier, make use of the wavelet transform to decompositionQRS wave band in the multi-scale, using modulus maxima method in combinationwith other methods to prevent undetection to calibration of R wave, finally, basing onthe relationship among the Q wave, S wave and T wave and R wave, we candetermine the starting point to determine the T wave, and the T wave data segment.(4)The part of T wave detection, Firstly this paper introduces the traditional andwidely application of spectrum analysis, and introduces its algorithm and simulationresults, the traditional detection algorithm of correlation analysis and correlationanalysis of quantitative research, analyzes the advantages and disadvantages of thetraditional the method of correlation analysis, and on this basis made improvements, so as to put forward an improved correlation analysis, introduced the corresponding Twhen band alignment, the method of using MATLAB to European electrocardiosignadatabase, MIT-BIH database ST-T and T-Wave Alternans Challenge database todo a lot of simulation experiments of electrocardiosigna, and presents someexperimental results to illustrate the feasibility of the method instead.In the end, the research direction, research methods are summarized andprospected, and improvements to the traditional method in theory is to improve itsreal-time performance, but for its judgement index, as well as the methods to improveafter the corresponding preprocessing part of ascension and improve remains to befurther research and study should be continued.