Diabetes risk early warning method and system based on depth auto-encoder

Abstract

The invention discloses a diabetes risk early warning method and system based on a depth auto-encoder. The method comprises the steps of obtaining and preprocessing a physical examination data set; dividing a label-free data set into a training set, a verification set and a test set, and generating a label of the depth auto-encoder; designing an encoder part and a decoder part of the depth auto-encoder, and training and evaluating the encoder part and the decoder part by using the training set, the verification set and the test set respectively; and predicting the diabetes risk of an individual to be detected by using the evaluated depth auto-encoder and the trained regression model. The invention further discloses diabetes risk early warning computer equipment based on the depth auto-encoder and a computer readable storage medium. The depth self-encoder is introduced, so that a small amount of label data can be fully utilized; more accurate diabetes early warning is carried out by training the deep auto-encoder in advance and combining a logistic regression model; and the method and system adapt to different diabetes types and symptoms, and are higher in practicability, generalization and expansibility.

Description

technical field [0001] The present invention relates to the field of deep learning and intelligent medical technology, in particular to a diabetes risk early warning method based on a deep autoencoder, a diabetes risk early warning system based on a deep autoencoder, computer equipment, and a computer-readable storage medium. Background technique [0002] In recent years, with the improvement of consumption level and the aggravation of population aging trend, the number of diabetic patients in my country is rising sharply. As a common chronic disease, diabetes can reduce the incidence rate and improve the quality of life of patients through scientific and effective intervention and prevention. Early warning and detection of diabetes will help reduce the national social welfare expenditure burden on diabetes and other chronic diseases and improve the quality of life of patients. However, the current level of awareness and prevention of diabetes among the national physical ex...

AI Technical Summary

Problems solved by technology

The main problem solved by the present invention is how to overcome the shortage of a large number of training labels in the traditional diabetes early warning model, and how to combine the deep self-encoder to intelligent medical treatment and diabetes risk early warning to provide a reliable and generalizable solution question

Images