ECG signal diagnosis method of ultimate convolutional neural network

Abstract

The invention discloses an ECG signal diagnosis method of an ultimate convolutional neural network. The method includes the following steps of S1, a data processing stage, wherein data comes from an MIT / BIH database, and a convolutional neural network of a three-layer structure is constructed; S2, a feature processing stage, wherein the convolutional neural network is utilized to conduct feature extraction on ECG waveforms in the database; S3, a parameter input stage, wherein feature output parameters of the convolutional neural network are input to an ultimate learning machine; S4, a trainingand learning stage, wherein the parameters of the ultimate learning machine are trained, and sample data of the ECG waveforms in the database is subjected to network training; S5, a signal diagnosisstage, wherein ECG signals are diagnosed through the trained ultimate learning machine. The purpose of diagnosing the ECG signals is achieved, the mode of manually monitoring the signals for diagnosing ECG is abandoned, a neural network mode is adopted for signal diagnoses, the effect is great, and the method is very worthy of popularization.

Description

technical field [0001] The invention relates to the technical field of computer program diagnosis signals, in particular to an ECG signal diagnosis method of extreme convolutional neural network. Background technique [0002] Cardiovascular disease caused by arrhythmia is a major health problem facing the world, and it can lead to temporary shock and even sudden death in patients. At present, accurate diagnosis and timely treatment are the most effective measures to deal with cardiovascular diseases. ECG is currently the most important means of detecting and diagnosing heart disease. However, the large amount of image information generated in the examination of diseases tends to make doctors tired, and the diagnosis accuracy is affected by subjective factors such as the professional ability and experience of the doctors. In this context, the use of machine learning methods to determine whether there is a problem with the heart or the specific type of heart disease has beco...

AI Technical Summary

Problems solved by technology

[0005] However, in the process of using the above-mentioned method of feature fusion-based ECG signal classification, there are still obvious defects: when classifying ECG signals, it is impossible to perform signal diagnosis on various waveforms of ECG, which makes it impossible to pass through the neural network well. The network diagnoses the signal. In the signal diagnosis, only manual monitoring and diagnosis can be used. Not only the diagnosis accuracy is poor, but also a waste of time and manpower, which is very inconvenient

Images