Disease diagnosis device and method based on medical images

The invention provides a disease diagnosis device and method based on medical images. The device comprises a feature extraction module used for performing feature extraction on an input medical image by using a convolutional neural network; the first diagnosis module is used for performing disease diagnosis reasoning based on the Bayesian network; the second diagnosis module is used for performing disease diagnosis and attribute category reasoning based on the graph convolutional network; and the diagnosis result fusion module is used for fusing the output of the first diagnosis module and the output of the second diagnosis module to obtain a final diagnosis result. According to the invention, disease diagnosis reasoning is carried out based on the Bayesian network and the graph convolution network, so that the disease intelligent diagnosis model can think like professional doctors and deduce diseases from imaging characterization of lesions, verifiable disease intelligent diagnosis is realized, professional knowledge of the doctors is fused, and the diagnosis accuracy is improved. Therefore, the accuracy, verifiability and generalization of automatic diagnosis are improved.