Automatic prediction and recognition method and system for echocardiogram based on artificial intelligence

Abstract

The invention discloses an automatic prediction and recognition method and system for an echocardiogram based on artificial intelligence. The method comprises the steps: acquiring a color Doppler video of at least one tangent plane of an echocardiogram of a detection object, extracting each video frame in the color Doppler video, inputting each video frame into a trained convolutional neural network to obtain an N-dimensional feature vector corresponding to each video frame, generating a weight corresponding to each video frame from the N-dimensional feature vector of each video frame throughan attention module, calculating the weighted sum of the N-dimensional feature vectors of each video frame by using the weight so as to obtain the overall feature representation of the color Doppler video, and based on the overall feature representation, calculating to obtain a prediction value containing the pre-recognized image feature. By means of the method, whether the to-be-recognized imagefeatures exist in the echocardiogram or not can be accurately predicted.

Description

technical field [0001] The present application relates to the technical field of intelligent recognition of medical video images, in particular to an artificial intelligence-based method and system for automatic prediction and recognition of echocardiography. Background technique [0002] Echocardiography is currently one of the most important methods for assessing cardiac structure and function. For the identification of the characteristics of heart valve regurgitation in echocardiography, doctors usually identify with the naked eye whether there is abnormal color blood flow, which is easily affected by the speed range and color gain, thus overestimating or underestimating the severity of the above features degree and is not suitable for precise assessment of the characteristics of cardiac valve regurgitation. In addition, there are vena contracta method, continuous Doppler and other methods that can quantitatively analyze the degree of cardiac valve regurgitation. However...

AI Technical Summary

Problems solved by technology

[0004] In recent years, although the use of artificial intelligence technology to process medical images is a research hotspot, more traditional machine learning algorithms such as decision trees, clustering, Bayesian classification, support vector machines, and EM are used. These algorithms do not use echocardiography. The dynamic video of the figure is only classified by randomly selected ultrasound images, which loses the motion information of the heart, and the color Doppler video contains the flow direction information of the blood flow in the heart, which is helpful to identify valve regurgitation, if such information is ignored It is easy to cause inaccurate classification, and the effect of practical application is general

Images