Neck lymph node image classification and recognition device and method

The invention relates to a neck lymph node image classification and recognition device and method, and the method comprises the steps: obtaining a plurality of groups of lymph node data images with determined classification types, and enabling each group of lymph node data images to comprise an ultrasonic image and a corresponding Doppler image which are generated through the detection of the samecolor ultrasonic machine; carrying out modal alignment on the ultrasonic images and the corresponding Doppler images in each group of lymph node data graphs to obtain a bimodal sample; constructing adeep learning model, and training the deep learning model by using the multiple groups of bimodal samples; wherein the deep learning model comprises two identical residual networks, a feature dimension splicing layer and an output network; and obtaining a lymph node image to be recognized, generating a corresponding input sample, and inputting the input sample into the trained deep learning modelfor classification and rec