Method and device for joint classification of ECG and cardiac shock signals based on neural network

Abstract

The invention provides a method for performing united classifying on an electrocardio signal and a cardiac vibration signal based on a neural network. The method comprises the following steps of obtaining the electrocardio signal and the cardiac vibration signal to be classified, and performing pretreatment, so as to realize the filtering and the noise elimination of the electrocardio signal and the filtering and the noise elimination of the cardiac vibration signal to be classified; extracting the characteristic waves of the electrocardio signal and the cardiac vibration signal to be classified, and converting the characteristic waves into a time-frequency diagram; transporting the time-frequency diagrams of the electrocardio signal and the cardiac vibration signal to be classified into a trained neural network for respective recognition, so as to obtain a recognition result; and connecting the results after recognition of the electrocardio signal and the cardiac vibration signal to be classified by a Concat method, and performing classification on the connected result through an Adaboost algorithm, wherein the neural network adopts a ResNet structure. According to the method disclosed by the invention, the one-dimensional electrocardio signal and the one-dimensional cardiac vibration signal are converted into the time-frequency diagram, and the neural network is combined with the Adaboost algorithm, so that the electrocardio signal and the cardiac vibration signal are effectively united for classification, the classifying dimensionality is increased, and the breakthrough of the classifying accuracy is achieved.

Description

technical field [0001] The invention belongs to the field of medical signal processing, and in particular relates to a neural network-based method and device for joint classification of electrocardiographic and cardiac shock signals. Background technique [0002] ECG monitoring technology uses the electrical excitation of the heart before each mechanical contraction to obtain heart status information. The common ECG detection technology usually obtains the ECG of the measurer through multiple electrode leads, so as to check for arrhythmia, ventricular and atrial hypertrophy, myocardial ischemia and other diseases. This technology is a relatively common heart detection technology in people's daily life. [0003] Cardiac shock monitoring technology uses the weak mechanical vibration caused by the heart pumping to obtain the state of the heart, which has high clinical research value. Compared with the electrocardiogram signal, the heart shock signal is composed of multiple ve...

AI Technical Summary

Problems solved by technology

However, this invention patent only uses a band-pass filter with a lower cut-off frequency lower than 1Hz and an upper cut-off frequency in the range of 100-250Hz for filtering, and the filtering effect on the signal is poor; at the same time, this invention patent requires additional use The microphone records the heart sound, and then divides the cardiac cycle according to the heart sound. This method is more complicated than directly using the heart shock signal to distinguish the cardiac cycle.

[0009] It can be seen that the existing patents are mainly for the separate classification of ECG signals and the classification of cardiac shock signals, and few of them combine the two to achieve classification.

At the same time, most of the current invention patents for ECG signal processing are based on one-dimensional ECG signal processing, which is not only limited to a single method, but is also limited by the cumbersome signal processing calculations and signal quality, which greatly reduces the accuracy of classification.

Images