Study On Automatic Identification Of Electrocardiogram P Wave For Atrial Fibrillation Assisted Diagnosis

Atrial Fibrillation(AF)is a common clinical disease of the heart.It is an arrhythmia with high risk,which can easily lead to multiple traumatic complications.Therefore,its identification becomes the main means to improve the survival condition of patients.because the AF has concealment,short-term electrocardiogram(ECG)is very difficult to find and it is a difficulty and hotspot for long-term ECG to accurate the data that a few seconds,dozens of seconds of AF in a few days of data.In AF diagnosis and AF minitoring,automatic identification of AF is extremely important for clinical treatment.The automatic identification of AF not only gives the cardiologist an accurate and reliable diagnosis,but also provides timely and effective advice for the individual.One of the important characteristics of AF recognition is to determine whether P waves exist in ECG.In this paper,a P wave automatic identification method based on gaussian model is proposed to solve the identification of AF.We have established different points according to the P wave form three types of P wave identify general Gaussian Model,including creating a type that contains a sleek P wave,pointed type P wave and bimodal type P wave Model group.Three types of P wave characteristics extracted data and by using Gaussian Mixture Model(GMM)training as P wave Gaussian Model group.Then the corresponding threshold range of Mahalanobis Distance(MD)training was used.In this paper,we selected the MIT-BIH Atrial Fibrillation Database as the experimental data,and the experimental results showed that the recognition accuracy rate of the smooth P wave,the sharp P wave,the double peak P wave and the linear data was 80.50%,84.36%,67% and 100% respectively.Because the ECG is a weak signal,which is highly affected by the environment,the proposed mechanism of P wave recognition will appear fake P wave phenomenon,which affect the diagnosis of AF.Therefore,the article puts forward the identification strategies of fake P wave based on BP neural network.The ture P wave data and fake P wave data are used as positive and negative examples to train BP neural networks.The experimental data were from the MIT-BIH Atrial Fibrillation Database,and the experimental results showed that the accuracy of fake P wave recognition was96.72%.