Realization method of atrial fibrillation detection based on deep convolutional neural network

A neural network and deep convolution technology, applied in the field of atrial fibrillation detection, can solve the problem of not finding the successful application of deep convolutional neural network, and achieve the effect of improving efficiency and accuracy

Active Publication Date: 2017-09-26
HARBIN INST OF TECH AT WEIHAI
View PDF3 Cites 54 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

However, due to the particularity of ECG signals, there has been no report on the successful a

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
  • Realization method of atrial fibrillation detection based on deep convolutional neural network
  • Realization method of atrial fibrillation detection based on deep convolutional neural network
  • Realization method of atrial fibrillation detection based on deep convolutional neural network

Examples

Experimental program
Comparison scheme
Effect test

Example Embodiment

[0023] Specific implementation mode 1: This embodiment provides a method for implementing atrial fibrillation detection based on a deep convolutional neural network, in order to make the original one-dimensional ECG signal applicable to the convolutional neural network structure in the form of processing two-dimensional data , The one-dimensional ECG signal needs to be appropriately transformed to meet the input signal requirements of the deep convolutional neural network. In addition, in the implementation method of atrial fibrillation detection based on a deep convolutional neural network provided by this embodiment, the deep convolutional neural network is constructed based on Caffe, one of the most popular deep learning frameworks. The specific implementation steps are as follows:

[0024] (1) The ECG acquisition equipment is used to collect the continuous ECG data of a single lead of the patient, and the doctor will mark each heartbeat of the ECG data whether there is atrial...

Example Embodiment

[0044] Embodiment 2: This embodiment provides a method for detecting atrial fibrillation based on static wavelet transform combined with deep convolutional neural network, such as figure 2 As shown, the specific steps are as follows:

[0045] Step (1): Read the records in the MIT-BIH atrial fibrillation database.

[0046] Step (2): Perform data segmentation on these ECG records. The duration of each data segment is 5 seconds. Since the sampling frequency of the MIT-BIH atrial fibrillation database is 250 Hz, each 5-second data segment contains 1250 Sampling point. According to the annotation file of the MIT-BIH atrial fibrillation database, a sample category label is set for each 5-second data segment. The basis for setting the label is: if the proportion of the number of heart beats of atrial fibrillation in each 5-second data segment in the entire data segment ≥ 50%, the data segment is marked as atrial fibrillation segment, otherwise it is non-atrial fibrillation segment.

[00...

Example Embodiment

[0058] Specific embodiment 3: This embodiment provides a method for detecting atrial fibrillation based on short-time Fourier transform combined with deep convolutional neural network, such as image 3 As shown, the specific steps are as follows:

[0059] Step (1): Read the records in the MIT-BIH atrial fibrillation database.

[0060] Step (2): Perform data segmentation on these ECG records. The duration of each data segment is 5 seconds. Since the sampling frequency of the MIT-BIH atrial fibrillation database is 250 Hz, each 5-second data segment contains 1250 Sampling point. According to the annotation file of the MIT-BIH atrial fibrillation database, a sample category label is set for each 5-second data segment. The basis for setting the label is: if the proportion of the number of heart beats of atrial fibrillation in each 5-second data segment in the entire data segment ≥ 50%, the data segment is marked as atrial fibrillation segment, otherwise it is non-atrial fibrillation s...

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

The invention discloses a realization method of atrial fibrillation based on a deep convolutional neural network. By the method, single-lead ECG data can be changed into a two-dimensional form through signal conversion to enable the same to be suitable for the deep convolutional neural network for processing two-dimensional data, so that automatic detection on atrial fibrillation is realized finally by automatically learning features and classifying the same by a machine. When the method is used for atrial fibrillation detection, P wave or R-R interval does not need to be detected, and artificial designing of features is not needed also, so that efficiency and accuracy in atrial fibrillation detection are improved greatly, wherein an atrial fibrillation detection method based on stationary wavelet transform combined with the deep convolutional neural network is 98.63% in accuracy, 98.79% in sensitivity and 97.87% in specificity, and an atrial fibrillation detection method based on short-time Fourier transform combined with the deep convolutional neural network is 98.29% in accuracy, 98.34% in sensitivity and 98.24% in specificity.

Description

technical field [0001] The invention belongs to the technical field of atrial fibrillation detection, and relates to an implementation method of atrial fibrillation signal detection in electrocardiographic signal identification, in particular to an implementation method of atrial fibrillation signal detection in electrocardiographic signals based on a machine learning algorithm. Background technique [0002] With the development of artificial intelligence technology, the detection of atrial fibrillation signals can no longer rely on the experience and judgment of professional doctors, and the use of machine learning algorithms in artificial intelligence technology allows machines to detect atrial fibrillation signals, which improves the efficiency of atrial fibrillation detection . Most of the traditional detection of atrial fibrillation signals based on machine learning algorithms needs to detect the P wave or R-R interval in the ECG signal first, and then use machine learn...

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
IPC IPC(8): G06F19/00G06K9/00G06N3/08
CPCG06N3/08G06F2218/06
Inventor 夏勇乌兰娜仁王宽全张恒贵
Owner HARBIN INST OF TECH AT WEIHAI
Who we serve
  • R&D Engineer
  • R&D Manager
  • IP Professional
Why Eureka
  • Industry Leading Data Capabilities
  • Powerful AI technology
  • Patent DNA Extraction
Social media
Try Eureka
PatSnap group products

Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic.

© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap