Automatic arrhythmia classifying method based on discriminant deep belief network

A deep belief network, arrhythmia technology, applied in the field of arrhythmia detection and classification, can solve problems such as validity impact

Active Publication Date: 2019-07-12
HARBIN UNIV OF SCI & TECH
View PDF5 Cites 19 Cited by
  • Summary
  • Abstract
  • Description
  • Claims
  • Application Information

AI Technical Summary

Problems solved by technology

[0004] The above feature extraction methods largely rely on the artificial design and

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
  • Automatic arrhythmia classifying method based on discriminant deep belief network
  • Automatic arrhythmia classifying method based on discriminant deep belief network
  • Automatic arrhythmia classifying method based on discriminant deep belief network

Examples

Experimental program
Comparison scheme
Effect test

specific Embodiment approach 1

[0080] A method for automatically classifying cardiac arrhythmias based on a discriminative deep belief network, the method comprising the steps of:

[0081] Step 1: ECG signal preprocessing;

[0082] Before the ECG signal is sent to the DDBNs network, it needs to be filtered, R peak location and normalization;

[0083] Step 2: DDBNs model construction;

[0084] (1) GRBM

[0085] DBNs is a graphical model that learns to extract deep representations of training data. It consists of stacked RBMs. RBMs are typical neural networks with visible and hidden layer interconnections. There is no connection between any two neurons in the same layer. An RBM with a binary input on the visible layer is a BB-RBM, while an RBM with a real-valued input on the visible layer is a GB-RBM;

[0086] (2) DRBM;

[0087] DRBM uses a single RBM with two sets of visible layers to train the joint density model. In addition to the unit A representing the input data, there is also a classification labe...

specific Embodiment approach 2

[0096] This embodiment is a further description of the discriminative deep belief network-based automatic arrhythmia classification method described in the first embodiment, and the first step includes the following process:

[0097] Firstly, the signal is decomposed by 9-scale wavelet, the detail coefficient of the first layer is 90-180 Hz and the approximation coefficient of the ninth layer is 0-0.35 Hz, and the other wavelet coefficients are reconstructed after adaptive wavelet threshold filtering to remove baseline drift and High frequency interference;

[0098] Then, determine the heart beat position by R peak positioning;

[0099] Finally, 256 sampling points were selected for the heart beat length including P wave and T wave, that is, 90 sampling points before and 165 sampling points after R peak were taken.

specific Embodiment approach 3

[0101] This embodiment is a further description of the arrhythmia automatic classification method based on the discriminative deep belief network described in the first embodiment, and the specific process of the step 2 GRBM is as follows:

[0102] The concrete process of described step two GRBM is:

[0103] DBNs is a graphical model that learns to extract deep representations of training data. It consists of stacked RBMs. RBMs are typical neural networks with visible and hidden layer interconnections. There is no connection between any two neurons in the same layer. The RBM with binary input on the visible layer is BB-RBM, and the RBM with real-valued input on the visible layer is GB-RBM. The energy functions of BB-RBM and GB-RBM are defined as formula (1) and formula ( 2) as shown:

[0104]

[0105]

[0106] Among them, θ 1 ={w ij ,b i ,c j}, θ 2 ={w ij ,b i ,c j ,σ j} represents the RBM parameters to be trained;

[0107] no v and n h are the number of visib...

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 an automatic arrhythmia classifying method based on a discriminant deep belief network, and belongs to the technical field of detecting and classifying of arrhythmia. The network is finely adjusted through back propagation, and automatic classifying of 6 kinds of arrhythmia types of the normal rhythm, the left bundle branch block, the right bundle branch block, the ventricular premature beat, the atrial premature beat and the pacing heartbeat is thus achieved. The automatic arrhythmia classifying method includes the steps that 1, ECG signals are pretreated; 2, a DDBNs model is built; 3, the DDBNs model is trained; 4, the DDBNs model conducts supervision and classification; 5, the ECG signals are sent to a DDBNs network, 256 sampling points x sent to the DDBNs modeland three-dimensional RR interphase characteristics r are normalized, the x is normalized, and a heartbeat normalized sample is obtained, and is sent to a first layer of the network. The automatic arrhythmia classifying method is applied to automatic classifying of arrhythmia.

Description

technical field [0001] The invention relates to the detection and classification of arrhythmia, in particular to an automatic classification method for arrhythmia based on a discriminative deep belief network. Background technique [0002] Physicians can diagnose arrhythmias by visually detecting short-term ECG signals, but they cannot rely solely on visual detection for long-term ECG signals recorded by dynamic electrocardiographs. With the development of computer science, the automatic classification and analysis technology of ECG signals has emerged as the times require, and now it plays an important role in the diagnosis and analysis of arrhythmia, and has become an important means of assisting clinical diagnosis of heart diseases. [0003] In the past ten years, several pattern recognition methods have been developed for the detection and classification of arrhythmias. [1-3] , mainly involves three steps: preprocessing, waveform detection and segmentation, feature extr...

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): A61B5/0402G06K9/62G06N3/04G06N3/08
CPCA61B5/7203A61B5/7264G06N3/084A61B5/318G06N3/045G06F18/24
Inventor 宋立新房奇
Owner HARBIN UNIV OF SCI & TECH
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