Artificial intelligence-based heart disease auxiliary detection method
A heart disease, auxiliary detection technology, applied in the directions of diagnostic recording/measurement, medical science, diagnosis, etc., to achieve the effect of reducing the time of diagnosis, improving the performance such as accuracy, and enriching the characteristics of dynamic pathological information
- Summary
- Abstract
- Description
- Claims
- Application Information
AI Technical Summary
Problems solved by technology
Method used
Image
Examples
Embodiment 1
[0042] Example 1. Sample set construction and sample data preprocessing
[0043] 1. Establishment of sample set
[0044] First, construct the sample set data for the convolutional neural network recognition model, the specific construction method is as follows:
[0045] 1.1 Composition of sample set
[0046] Include n clinically known heart healthy individuals (n>500) and m clinically known certain heart disease individuals (m>1000) as the sample population; collect image data of the sample population related to the specific heart disease as the sample set data; wherein, the image data of the sample population includes but not limited to vector cardiogram and electrocardiogram.
[0047] 1.2 Setting of sample label
[0048] The index data of the gold standard index of heart disease is used, and the expert consensus in the industry and the diagnosis information of doctors in different tertiary hospitals are used to jointly determine the disease label of the sample data. Lab...
Embodiment 2
[0055] Example 2. Acquisition of image data of various pathological features related to attributes of a specific cardiac disease
[0056] After obtaining the preprocessed sample set data in Embodiment 1, the pathological feature image data of a specific cardiac disease is obtained. The pathological feature image data of a specific heart disease includes an electrocardiogram and a vector cardiogram, etc.; the specific operation process is carried out according to the following steps:
[0057] Acquisition of vector cardiogram: collect labeled cardiac electrical signal data, perform preprocessing such as median filtering and wavelet transform filtering on the collected cardiac electrical signal data, and then convert to obtain vector cardiogram (VCG). The conversion method adopts KorsJ.A. et al. published in 1990 in European Heart Journal Journal of the 11(12):1083 paper describes the method and parameters to simultaneously obtain multidimensional X(t) , Y(t) and Z(t) ECG...
Embodiment 3
[0059] Example 3. Constructing and optimizing the convolutional neural network model on the preprocessed image data of pathological characteristics of cardiac diseases
[0060] 1. Constructing a Convolutional Neural Network Recognition Model for Specific Cardiac Diseases
[0061] Using the image data of the pathological features of the specific heart disease obtained in Example 2 as input data, machine learning is carried out to construct a convolutional neural network model adapted to the specific heart disease, and to realize the correlation between the quantitative index data of each pathological feature and the specific heart disease. One-to-one correspondence between attributes; specifically: building a convolutional neural network recognition model for electrocardiograms and electrocardiograms. A convolutional neural network recognition model of electrocardiogram or electrocardiogram, comprising at least one input layer, at least one hidden layer and at least one output ...
PUM
Abstract
Description
Claims
Application Information
- R&D Engineer
- R&D Manager
- IP Professional
- Industry Leading Data Capabilities
- Powerful AI technology
- Patent DNA Extraction
Browse by: Latest US Patents, China's latest patents, Technical Efficacy Thesaurus, Application Domain, Technology Topic, Popular Technical Reports.
© 2024 PatSnap. All rights reserved.Legal|Privacy policy|Modern Slavery Act Transparency Statement|Sitemap|About US| Contact US: help@patsnap.com