Method and device for solving irregular problem of medical time sequence data

The invention relates to the technical field of medical data processing, in particular to a method and device for solving the problem of irregular medical time sequence data. The method comprises the following steps: encoding input spatio-temporal data to obtain an input sequence; the method comprises the steps of obtaining an input sequence, performing feature recognition and feature extraction on the input sequence to obtain a node feature vector, generating a preliminary filling result according to an MPNN model, and optimizing data points based on the preliminary filling result to generate a secondary filling result. According to the method, the neural network model is constructed, the data are coded and decoded in the time dimension and the same operation can be carried out in the space dimension by effectively utilizing the internal relevance of the time sequence data aiming at the input irregular data, and in the filling of the first stage, the time sequence data is not required to be coded and decoded