Alzheimer's disease progress prediction method

Abstract

The invention discloses an Alzheimer's disease progress prediction method, a patient is measured in a fixed time interval to obtain multi-view data, and the Alzheimer's disease progress prediction method comprises the following steps: preprocessing the multi-view data; processing the multi-view data by using a multi-view fusion neural network to obtain a hidden representation matrix shared by each view; introducing a minimum gating unit into the implicit representation matrix so as to fill missing multi-view data and predict score data of a future time point; and carrying out cooperative training on the multi-view fusion neural network and the minimum gating unit by using the multi-view data so as to predict the disease development process. According to the method, the hidden representation matrix shared among the views is obtained for the multi-view data of each time point through the multi-view fusion neural network, meanwhile, the data, obtained through prediction of the minimum gating unit, of the next time point is used for filling missing data, and score data prediction of any time point in the future is conducted through the minimum gating unit.

Description

technical field [0001] The invention relates to a method for predicting the progress of Alzheimer's disease, which belongs to the field of medical artificial intelligence. Background technique [0002] Alzheimer's disease (AD) is an irreversible, progressive neurodegenerative chronic disease, which will gradually destroy the patient's memory and cognitive ability, and eventually lead to the patient's death. At present, there are more than 50 million AD patients in the world, and it is estimated that the number of patients will reach 114 million by 2050. AD not only brings endless pain and mental pressure to the patients themselves, but also to the patients' families and the whole society. a huge economic burden. Unfortunately, AD disease can only be controlled but not completely cured. It usually lasts for a long time and develops slowly. Therefore, it is particularly important to carry out early detection in the pre-symptomatic stage of patients and implement interventio...

AI Technical Summary

Problems solved by technology

[0007] The purpose of the present invention is to provide a method for predicting the progress of Alzheimer's disease, which can effectively use multi-view data and solve the problem of missing longitudinal data of multi-view, and fully tap and utilize the potential correlation between time series of multi-view data information and build a flexible multi-point prediction model to predict the longitudinal cognitive score trajectory, which can flexibly process incomplete multi-view time series data and use it to predict the score data at any time point in the future, and integrate the multi-view neural network and the minimum gate The control unit is integrated into a unified framework for collaborative training to help the network learn the optimal feature representation and model parameters, and then realize the prediction of diseases

Images