Respiratory disease patient person-time dynamic prediction method based on neural network

Abstract

The invention discloses respiratory disease patient person-time dynamic prediction method based on a neural network. According to the method, the distribution of outpatients and inpatients of respiratory diseases is extracted from medical records, and regional air pollutant concentration, regional meteorological data, holidays, day of the week effect and regional socio-economic level and other information are integrated. An LSTM-based neural network model is applied to implement the method. The method comprises the follow steps: step 1: integrating multi-source data; step 2: data preprocessing; step 3: constructing the LSTM neural network model; step 4: model verification and optimization; and step 5: incremental learning and dynamic prediction. The scientific basis for solving the problems of optimizing the allocation of regional medical and health resources and studying the impact of air pollution on the burden of disease can be provided through dynamic prediction of the respiratorydisease patient person-time with high accuracy.

Description

technical field [0001] The present invention relates to the field of prediction method of patients, data preprocessing technology and artificial neural network technology, in particular to a dynamic prediction method of patients of respiratory diseases based on neural network. Background technique [0002] With the development of industrialization and the acceleration of urbanization, the problem of air pollution has become increasingly serious, resulting in frequent occurrence of smog in my country in recent years. Air pollution poses a great threat to human health and has aroused widespread concern from people from all walks of life at home and abroad. People urgently hope that the government and relevant departments will give timely forecasts of relevant sensitive diseases based on weather and air pollution conditions, so that protective measures can be taken in advance to minimize the adverse effects of air pollution on human health. [0003] The impact of air polluti...

AI Technical Summary

Problems solved by technology

[0006] Therefore, the research purpose of the present invention is to combine the main pollutants (PM2.5, PM10, NO2, SO2, CO, O3) concentration data of air pollution, meteorological data such as daily average temperature and relative humidity every day, regional socio-economic level Data, regional disease hospitalization data and outpatient and emergency department visit data, provide a prediction method using Long Short-Term Memory (LSTM) to construct respiratory disease patients, so that with high accuracy Dynamically predict the number of patients with respiratory diseases in the region, and provide a scientific basis for solving problems such as optimizing the allocation of regional medical and health resources and studying the impact of air pollution on the burden of disease

Images