Design And Implementation Of Arrhythmia Assistant Diagnosis System

In recent years,due to changes in eating habits and the aging of the population,the incidence of cardiovascular disease has been increasing year by year,and it has become the disease with the highest incidence of urban and rural residents in China.Cardiovascular disease has the characteristics of concealment,acute onset,and high mortality.Early detection is the key to treatment,and electrocardiogram is the gold standard for diagnosing arrhythmia.Arrhythmia can be diagnosed by electrocardiogram.The arrhythmia assisted diagnosis system can classify the heartbeat classification at the same time as the ECG signal is detected in real time,providing a reference for the doctor to diagnose,and then make up the gap between the doctors’ level.At present,the common arrhythmia assisted diagnosis system still has many deficiencies,such as the high computational complexity of the ECG signal detection algorithm and the poor detection effect on abnormal ECG signals;the single feature extraction of the arrhythmia classification algorithm and the poor classification effect;the malfunction-assisted diagnosis system lacks the real-time diagnosis function of ECG signals and other problems.In view of the above problems,this paper focuses on ECG signal detection algorithm and arrhythmia classification algorithm,and designs and implements a portable arrhythmia assisted diagnosis system.The research results of this paper are as follows:1.Based on the real-time requirements of the system,combined with Shannon energy and improved adaptive threshold method,a fast QRS wave detection algorithm is proposed.First of all,the signal is denoised and normalized pre-processed;then the Shannon energy of the processed signal is calculated,the improved adaptive threshold method is used for positioning,and the signal after QRS wave enhancement is used to modify the positioning result;finally,a back-check is added Mechanism to improve the accuracy and stability of QRS wave detection.The performance of the proposed algorithm is evaluated based on the data of the MIT-BIH arrhythmia database.The analysis results show that the QRS wave detection algorithm in this paper has high P waves and T waves in the signal,irregular heart rhythm,and the signal is seriously interfered by noise The position of the QRS wave can still be accurately located,and the sensitivity,positive detection degree and accuracy rate of 109494 heart beats in the database have reached 99.88% respectively.99.85% and 99.73%.While ensuring the accuracy of detection,the algorithm has low computational complexity,which is beneficial to the rapid QRS wave detection of the arrhythmia assisted diagnosis system.2.In order to improve the accuracy of arrhythmia classification of the arrhythmia assisted diagnosis system,an arrhythmia classification algorithm based on fusion features and random forest is proposed.In this paper,the signal is first denoised;then medical features,statistical features,and morphological features are extracted for each heartbeat,and the three features are cascaded to enhance the ability to express features.Among them,the medical feature is the RR interval,which can reflect the change of heart rate;the statistical features include the kurtosis coefficient,skewness coefficient,and standard deviation,which can reflect the change trend of the amplitude of the ECG signal;The packet decomposition coefficient can reflect the morphological changes of the ECG signal.Finally,the accuracy of the arrhythmia classification of the three classifiers of support vector machine,BP neural network and random forest is compared.The five types of arrhythmia data in the MIT-BIH arrhythmia database were used to test in two classification modes: intra-patient and inter-patient.Comparing the results of the three classifiers,the classification accuracy based on random forest is the highest,and the model training and the classification time of a single heartbeat are shorter.In the intra-patient classification mode,a 10-fold cross-validation method is used for grouping,and the obtained classification accuracy,specificity and sensitivity are 99.08%,99.00% and 89.31% respectively;in the inter-patient classification mode,22 sets of data are used as the training set,Using another 22 sets of data as the test set,the classification accuracy and specificity obtained were 92.31% and 89.98%,respectively.This algorithm achieves a high classification accuracy in the classification of common arrhythmia,and provides arrhythmia classification method for the auxiliary diagnosis system of arrhythmia.3.Established a portable diagnostic system for arrhythmia and tested the system function.This paper establishes a portable arrhythmia assisted diagnosis system through system hardware selection,circuit construction and program design.The AD8232 sensor is selected as the signal acquisition module,the STM32 minimum system is used as the signal sampling module,and the Raspberry Pi 3b is used as the signal processing module and combined with the display screen to complete the function of displaying results and user interaction.In the program design part,the design of digital filter,QRS wave detection algorithm,arrhythmia classification algorithm is completed,and the user interaction interface is designed to encapsulate the algorithm.System tests show that the system in this paper can achieve the functions of ECG signal acquisition,real-time signal display,heart rate calculation,heartbeat classification and signal storage,and meet the requirements of portable arrhythmia auxiliary diagnosis system.