Method, system and diagnostic device for constructing a cardiovascular disease diagnostic model

Abstract

The invention discloses a construction method and system of a cardiovascular disease diagnosis model and the diagnosis model. The method includes: constructing a labeled side face data set; training a cascade classifier to obtain an ear detection model; respectively using VGG , GoogleNet and ResNet neural network models to extract the features of the ear object; use the spatial pyramid to integrate the features of the ear features extracted by the neural network model, and obtain the depth heterogeneous feature map of each neural network model; Feature preprocessing; training to obtain the SVM classifier model; integrating the SVM classifier model and the three trained neural network models through Bagging learning to obtain a cardiovascular disease diagnosis model. The cardiovascular disease diagnostic model constructed by the present invention can comprehensively and scientifically diagnose and predict cardiovascular diseases with high precision, and can be widely used in the field of automatic processing of medical data.

Description

technical field [0001] The invention relates to the technical field of computer software, in particular to a method and system for constructing a cardiovascular disease diagnosis model and the diagnosis model. Background technique [0002] Most of the existing artificial intelligence models for cardiovascular disease detection are based on X-ray photos or electrocardiograms of the patient's heart. Obtaining these pictures requires a relatively cumbersome process and the support of a large number of professional equipment, and High cost and time-consuming. The existing schemes for disease diagnosis through face photos usually only extract the features of certain areas of the face, such as forehead, nose, etc., and then use traditional machine learning algorithms to classify, so as to realize diagnosis based on the classification results. The relationship between the characteristics of these locations and specific diseases lacks scientific basis, so the accuracy rate is low. ...

AI Technical Summary

Problems solved by technology

Traditional technology uses machine learning methods to extract ear features, but the features extracted by traditional machine learning methods are relatively single, usually only simple features such as color and texture can be extracted, and the extraction of these features lacks guidance, and the final features are often different from the Cardiovascular disease association is weak, so the diagnosis of cardiovascular disease cannot be made accurately

Images