Font Size: a A A

Arrhythmia Secondary Diagnosis

Posted on:2011-01-21Degree:MasterType:Thesis
Country:ChinaCandidate:X L ChenFull Text:PDF
GTID:2204360305993550Subject:Biomedical engineering
Abstract/Summary:PDF Full Text Request
In recent years, cardiovascular disease has been one of the three killers which threaten human's health, and the amount of heart disease suffers increases by degrees year after year. Arrhythmia is extremely common in clinic, so timely and accurate detection of arrhythmia has important clinical significance for prevention and cure of many kinds of heart diseases. This paper researches arrhythmia assistant diagnosis method, which can detect several kinds of arrhythmia and provide good assistance for efficient and accurate diagnosis of arrhythmia, aiming at technical problems of arrhythmia automatic analysis and combining much theoretics.Firstly, research the arithmetic of detecting ECG waves on the base of several main methods of detecting QRS complex waves. Implement an arithmetic of detecting ECG waves, using wavelet transform combined with the time-domain analysis. Evaluate the arithmetic through MIT-BIH, and get a good result, where the detection rate of R wave is 99.41%, the characteristics of P wave and T wave are accurately located.Secondly, extract effective feature parameters after detection, and detect arrhythmia using logical branch method. Both the choice of feature parameters and the formulation of arrhythmia discrimination rules depend on experience, and have subjective factor. So this paper uses attribute reduction based on rough set to choose feature parameters, and formulate arrhythmia discrimination rules based on theā…”lead referring to the clinical diagnostic criteria and experts. The rules applied to MIT-BIH, obtain a good result.Thirdly, get the membership of abnormal heart beats using fuzzy neural network for the shortage of logical branch method, providing more assistance for doctors to diagnose.Fourthly, in terms of software programming, use C# mixed with MATLAB:achieve the pretreatment of ECG, detection of waves, extraction of parameters, logical discrimination and membership using several toolboxes of Matlab; design user interface and general framework using C#, develop database using SQL Server, and complete the software design of arrhythmia assistant diagnosis system. It is evaluated through MIT-BIH and a good result is obtained.
Keywords/Search Tags:ECG, arrhythmia, wavelet transform, rough set, fuzzy neural network
PDF Full Text Request
Related items