Blood glucose change prediction method and equipment based on multi-modal data

The embodiment of the invention discloses a blood glucose change prediction method and device based on multi-modal data, and relates to the technical field of blood glucose prediction.The method comprises the steps that multi-modal blood glucose data of a patient is obtained, the multi-modal blood glucose data comprises any one or more of patient personalized information, patient physical measurement data, patient physiological parameter measurement data, patient blood glucose sensing signal data and patient daily activity event data; performing feature extraction on the multi-modal blood glucose data through a pre-trained stack noise reduction auto-encoder to determine a feature vector set corresponding to the multi-modal blood glucose data; and according to a pre-trained multi-modal fusion model, performing feature fusion on the feature vectors of the multiple modals in the feature vector set to output a blood glucose change range prediction result of the patient. The problem that manual feature extraction