Non-invasive cardiac infarction classification model construction method based on stack type self-encoder and support vector machine

Abstract

The invention discloses a non-invasive cardiac infarction classification model construction method based on a stack type self-encoder and a support vector machine. At a training stage, movement information of a heart is obtained by using a collected cardiac magnetic resonance image; a self-encoder is trained by using an image block based on fusion of image information and movement information as an input and a corresponding infarction situation as a tag; inputted data are preprocessed by using a denoising self-encoder; all variable factors of the upper layer are utilized during the high-dimensional information learning process and thus deep features of the inputted data are learned; and then with a support vector machine, sample classification is carried out by using the learned deep features as the input and a corresponding tag. According to the invention, cardiac infarction classification prediction is realized from the perspective of data driving, so that a problem of time and effort wasting caused by infarction prediction with contrast agent injection clinically is solved.

Description

technical field [0001] The invention belongs to the technical field of medical image analysis, and in particular relates to a method for constructing a non-invasive cardiac infarction classification model based on a stacked autoencoder and a support vector machine. Background technique [0002] Myocardial infarction (MI) is a common clinical cardiovascular system disease and a critical type of coronary heart disease with a high mortality rate. Myocardial necrosis will further cause cardiac remodeling, leading to arrhythmia and heart failure. Cardiac remodeling includes ventricular remodeling, vascular remodeling, nerve remodeling, and electrical remodeling. Changes, increased cardiac load, decreased compliance, decreased subintimal myocardial perfusion, increased myocardial oxygen consumption, disturbance in the initiation of compensatory mechanisms, myocardial electrical-mechanical desynchronization; morphologically manifested as cardiomyocyte hypertrophy, cardiomyocyte apo...

AI Technical Summary

Problems solved by technology

[0003] Previously, many methods for the detection of myocardial infarction have been proposed, such as JW van Oorschot et al. have demonstrated the feasibility of detecting patients with chronic myocardial infarction without the use of exogenous contrast agents; however, the sensitivity and specificity of this method low accuracy and cannot provide sufficient information on myocardial tissue characteristics; Asha et al. developed an intelligent heart disease prediction system using three classifier decision trees, naive Bayesian and neural networks to predict heart disease; Anbarasi M et al. proposed enhanced prediction of heart disease using genetic algorithms for feature subset selection, down from the original thirteen attributes for predicting heart disease to six attributes; however, none of them directly and accurately predicted the location and area of ​​MI

Carloz Ordonez et al. propose four constraints to reduce the number of rules: item filtering, attribute grouping, maximum itemset size, and pre / following rule filtering; however, when applying association rules to medical datasets, a large number of rules will be generated , most of these rules are medically trivial, and the time required to find them is impractical

Images