Electrocardiosignal QRS wave group identification method based on deep learning

Abstract

The invention brings forward an electrocardiosignal QRS wave group identification method based on deep learning, comprising the following steps: firstly setting a data preprocessing model; normalizingdata sampling rate to a preset frequency threshold, carrying out equal-length splitting on the normalized data to obtain data segments with length d; preprocessing the tagged sample data through thedata preprocessing model, and training according to the preprocessed sample data to obtain a prediction model that outputs probability y containing a QRS wave group in the data segments; preprocessingthe electrocardiosignal data according to the data preprocessing model, and inputting the data segments obtained by preprocessing into the prediction model to obtain probability corresponding to eachdata segment; then selecting data segments corresponding to probability which is greater than the preset frequency threshold to form a QRS wave group. The method is dependent on data and has greaterpotential and value for use in today's rapid development of medical informatization and accumulation of large amounts of data.

Description

technical field [0001] The present invention relates to the field of medical technology, in particular to a method for identifying QRS complexes of electrocardiographic signals based on deep learning. Background technique [0002] Electrocardiogram (ECG) records the electrophysiological signals of heart beating. Each beat can be divided into P wave, QRS complex (QRS-complex), T wave, etc., which correspond to atrial depolarization activity and ventricular depolarization respectively. Atrial repolarization activity, ventricular repolarization activity. Among them, the QRS wave group is the most obvious characteristic band of the electrophysiological signal of heart beating, which reflects the myocardial activity when the ventricle contracts and ejects blood. [0003] The identification of the QRS complex is the first step in ECG Interpretation. Only by correctly identifying the position of the QRS complex can we: (1) calculate the duration of each heart activity, and then ca...

AI Technical Summary

Problems solved by technology

In the above cases, it is easy to mistake P waves or T waves for QRS complexes using pattern recognition methods

[0008] susceptible to noise

During the acquisition process, different types of noise such as limb movement, electrical interference from nearby equipment, and muscle electrical interference will be encountered. These noises superimposed on the QRS wave group will change the shape and reduce the accuracy of the pattern recognition method.

Images