A method, device and application of medical examination data completion based on side information

Abstract

The invention discloses a physical examination data supplementing method based on edge information. The method comprises the steps of (1), constructing and supplementing a physical examination-diseasematrix, a pathogenic factor-disease matrix, a pathogenic factor-physical examination matrix according to the edge information; (2), establishing coding-decoding networks D2F Net, D2C Net and F2C Netbetween two random matrixes; (3), performed combined training on the D2F Net, D2C Net and F2C Net, wherein after training is finished, the pathogenic factor-physical examination matrix is supplemented; and (4), inputting the to-be-supplemented physical examination-disease matrix into the D2F Net, D2C Net, and by means of the supplemented pathogenic factor-disease matrix, the pathogenic factor-physical examination matrix and the F2C Net, supplementing the physical examination-disease matrix through calculation. The invention further discloses a physical examination data supplementing device based on the edge information, wherein the physical examination data and the disease result can be supplemented according to existing information.

Description

technical field [0001] The invention belongs to the field of data statistics and artificial intelligence, and in particular relates to a side information-based medical examination data complement method, device and application. Background technique [0002] The traditional physical examination plan is to carry out disease screening through a series of physical examinations: according to the needs of different diseases, physical examinations of relevant physiological characteristics are carried out under the arrangement and suggestion of doctors or medical manuals, and then the doctors pass the relevant physiological characteristics physical examination results to patients. Diagnose possible diseases. Due to the large number of physical examination items, different hospitals, doctors and times have different examination methods, resulting in a variety of physical examination items that cannot be unified, resulting in a waste of relevant medical resources and making patients f...

AI Technical Summary

Problems solved by technology

The side information method is still a branch of traditional machine learning, and there are not enough attempts to combine it with artificial neural networks and deep learning.

[0007] In the medical field, data loss is so serious, and labeled data is rare, but matrix completion methods are rarely applied

Images