Unlock instant, AI-driven research and patent intelligence for your innovation.

Training method and model of convolutional neural network based on improved sample distribution

A technology of convolutional neural network and sample distribution, which is applied in the field of artificial neural network in the field of heartbeat type screening, can solve the problems of misjudgment of heartbeat type, time-consuming and labor-intensive, etc., and achieve the effect of reducing misjudgment

Active Publication Date: 2021-09-03
SHANGHAI SID MEDICAL CO LTD
View PDF3 Cites 2 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

Screening out all premature ventricular beats from a long period of ECG signals is extremely important for judging the patient's condition, but it is time-consuming and labor-intensive, so it is of great significance to use a computer to screen out premature ventricular beats in ECG signals
However, since there are many positions in the heart where electrical shocks can be emitted, the shapes of the heart beats displayed are different when the electric shock positions are different, and there are many disturbances. It is almost impossible to train the deep learning model with the pattern and data of heart beats, which can easily lead to misjudgment of heart beat types.

Method used

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
View more

Image

Smart Image Click on the blue labels to locate them in the text.
Viewing Examples
Smart Image
  • Training method and model of convolutional neural network based on improved sample distribution
  • Training method and model of convolutional neural network based on improved sample distribution
  • Training method and model of convolutional neural network based on improved sample distribution

Examples

Experimental program
Comparison scheme
Effect test

Embodiment 1

[0030] This embodiment provides a training method based on an improved convolutional neural network based on sample distribution, such as figure 1 shown, including the following steps:

[0031] Step S31: Collect sufficient number of lead II beat signals in the marked 12-lead ECG signals to form an initial data set. The lead II beat signals include premature ventricular beat signals and non-ventricular premature beat signals each accounting for half of the total number , the label of the premature ventricular beat signal is set to a, and the label of the non-ventricular premature beat signal is set to b;

[0032] Step S32: Perform statistics on all heart beat data in the collected data set, count the mean value, maximum value, minimum value, peak value and kurtosis value of each heart beat data to form an average value array AVG={a1, a2, a3...aN }, maximum value array MAX={m1,m2,m3...mN}, minimum value array MIN={c1,c2,...cN}, peak value array F={f1,f2,...fN} and kurtosis value...

Embodiment 2

[0040] This embodiment provides an improved convolutional neural network model based on sample distribution, which is obtained by training the convolutional neural network model based on improved sample distribution described in Embodiment 1.

Embodiment 3

[0042] This embodiment provides a method for using an improved convolutional neural network model based on sample distribution, including:

[0043] Step S1: collect the heartbeat in the heartbeat signal of lead II in the 12-lead ECG signal of unknown type, and calculate the average value of the heartbeat x avg , the maximum value x MAX , minimum value x MIN , peak x F , and the kurtosis value x Q ;

[0044] Step S2: Judgment The number of established in;

[0045] Step S3: If there are 3 or more established in step S2, then input to the improved convolutional neural network model based on the sample distribution of embodiment 2 to determine the type of heart beat, otherwise it is considered that the unknown type of heart beat is "undeterminable type";

[0046] Step S4: When the output value of the convolutional neural network model based on the sample distribution is close to a, it is considered that the unknown type of beat is a premature ventricular beat; when the ...

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

PUM

No PUM Login to View More

Abstract

This application relates to a training method and model of a convolutional neural network based on sample distribution improvement provided by this application, by calculating the average value x of heart beats in advance avg , the maximum value x MAX , minimum value x MIN , peak x F , and the kurtosis value x Q , whether it meets the conditions to screen the samples, and when the number of established ones is less than 3, it is considered that the training data used by the model does not conform to the sample distribution, and these heartbeat signals that do not meet the requirements are eliminated, so as to reduce the misjudgment of the heartbeat type.

Description

technical field [0001] The application belongs to the technical field of heartbeat type screening by artificial neural network, and in particular relates to a training method and model of a convolutional neural network improved based on sample distribution. Background technique [0002] The ECG signal is a comprehensive reflection of the electrical activity of numerous cardiomyocytes in the heart. Under normal circumstances, the cardiomyocytes of the sinoatrial node, atrium, and ventricle are depolarized in turn, forming a heart beat that includes a signal segment of P wave, QRS wave, and T wave, and multiple heart beats are combined to form an ECG signal. Premature ventricular beat refers to the beat formed by depolarization of the ventricle caused by an electrical impulse sent out in advance by any part of the ventricle or an ectopic rhythm point in the ventricular septum before the impulse of the sinoatrial node reaches the ventricle. Screening out all premature ventricu...

Claims

the structure of the environmentally friendly knitted fabric provided by the present invention; figure 2 Flow chart of the yarn wrapping machine for environmentally friendly knitted fabrics and storage devices; image 3 Is the parameter map of the yarn covering machine
Login to View More

Application Information

Patent Timeline
no application Login to View More
Patent Type & Authority Patents(China)
IPC IPC(8): A61B5/349
CPCA61B5/7225A61B5/7267A61B5/725A61B5/364A61B5/366G16H50/20G16H50/70G06F2218/16G06F2218/14G06F18/2414G06F18/2148
Inventor 朱俊江王雨轩陈红岩朱志超张顺宇
Owner SHANGHAI SID MEDICAL CO LTD