Real-time heart disease screening method based on recurrent neural network

The invention provides a real-time heart disease screening method based on a recurrent neural network. A standard electrocardiogram paper image, a standard heart sound drawing image, an image data extraction model, a BERT data correlation detection model, an electrocardiogram diagnosis model, a heart sound diagnosis model and a fusion output model are included. The invention relates to the field of computer vision and heart disease diagnosis, in particular to a real-time heart disease screening method based on a recurrent neural network. The method has the advantages that the electrocardio diagnosis model, the heart sound diagnosis model and the fusion output model comprehensively consider electrocardio data and heart sound data, and grading diagnosis results can be given according to the severity of heart disease conditions; the dimension of the data is rapidly reduced by using the advantage of Transform parallel computing, and then accurate diagnosis is performed by using an RNN recurrent neural network; a