The Prediction Research Of Outpatients Amount Of Respiratory Medicine Department Based On SARIMA-RBF Combination Model

The hospital outpatient amount is one of the most direct and critical indicators reflecting the basic operation of medical institutions.Not only can it truly reflect the scale,quality and level of medical institutions,it also provides the medical administration and management with important reference so that reasonable allocations of medical resources could be made accordingly and the efficiency of medical staff can be improved.Therefore,it is of great significance to comprehend and predict the changing trend of outpatient visiting amount in an effective and accurate way within the field of medical information management.As the reform of medical informalization deepens and develops forward,the statistical prediction of outpatient amount plays a core role in the daily management and decision-making of medical institutions.With the worsen air quality and severe environment pollution,respiratory diseases occur frequently which has become a pressing worry as well as an urgent medical problem to be solved.The department of respiratory medicine is one of the core departments within the hospital,focusing on providing the treatment to bronchitis,tracheitis,asthma and other respiratory diseases.An inescapable fact remains that its daily consultation visiting amount is closely related to air quality.Currently,the majority of the prediction problems surrounding medical information management still remain in the initial stage of manual judgment on the basis of medical staff’s personal working experience,which indicates that there is a lack of accurate and reliable auxiliary prediction methods for effective medical forecasting.This thesis takes Wuhan,the city where the author studies and lives,as the research sample city.The factor of air quality is set as one of the key factors affecting the outpatient visiting quantity of respiratory medicine department.The outpatient data of respiratory medicine department of a hospital in Wuhan and the air pollution data of Wuhan from 2012 to 2016 are selected and put to use.In terms of the prediction problem of outpatient visiting amount in respiratory medicine department,SARIMA-RBF combination model is used to predict the changing trend of outpatient visits of the forthcoming month.Meanwhile,seasonal factors(Week)and random factors(Air Quality)are taken into account to complete the construction of the model,which shall provide scientific basis for the daily work arrangement and decision-making of the respiratory department of the hospital.The main innovation part of this research lies in the following three aspects: For starters,the Index Selection;This thesis proposes an air pollution index according to the characteristics of the patients of the respiratory medicine department,and takes the current air pollution level as a critical indicator affecting the visiting number of patients in the respiratory medicine department of approaching days.Secondly,the Selection of Time Granularity;Most of the previous studies relating to the prediction of outpatient visiting has adopted “Month” as the statistical dimension,which could only reflect the general trend of recent years.Nevertheless,this thesis refines and adopts “Day” as the statistical dimension instead.Last but not the least,the Optimization of Combination Model;much of the researches have been done at home and abroad considering the combination of predicting models,which includes the proposal of SARIMA-RBF combination model.However,previous research only took the AR term and predicted value in SARIMA as input to complete the training of RBF.In this thesis,the air pollution index is extracted from the actual data and used as the input variable of RBF.The combination of the two models is innovatively optimized to improve the prediction effect of respiratory outpatient visiting amount.The main work of this thesis is elaborated as follows:The developing history and status quo of current prediction problems within the medical field is systematically and comprehensively analyzed,which fully illustrates the research significance and extensibility of this thesis.The HIS(Hospital Information System)of the hospital and the impact of air pollution on human health is thoroughly studied and the positive correlation between air pollution and outpatient visiting number in respiratory medicine is comprehended.The commonly used SARIMA-RBF combination model is carefully adjusted,and the air factor is served as the input value of RBF in residual fitting,so that the whole structure of the combination model shall be optimized,more specific,thus the prediction accuracy can be improved.Through the analysis and the comparison of the prediction results between the single prediction model and combination model,this thesis finds that the combination model can not only give full play to the advantages of the traditional time series model in linear part fitting,also the neural network model can be used to better fit the nonlinear interference factors,therefore the prediction effect outweighs that of single model.Due to the Limitation of time,region and capability,the research content of this thesis still has room to be improved.For example,other than the factor of air quality,subsequent investigations of this matter should explore how other influencing factors can be added so as to further optimize the prediction effect of the combination model.