The Design And Implementation Of The Elderly Health Assessment System

As aging process speeds up,China is facing with the increasing social pressures caused by elderly health issues.The existing health assessment systems often use fixed evaluation rules and assessment models to assess the health status of individual,and that makes the elderly health assessment model cannot be constructed dynamically and conveniently by medics.Besides,the existing elderly health assessment institutions is not perfect and lacks control over the standard.Therefore,it is particularly significant to improve the health assessment institutions for the elderly and to build a visual health assessment model.In this paper,I will put forward my ideas to design and implement an elderly health assessment system,to improve the elderly health assessment institutions,to research the machine learning methods applicable to the elderly health assessment model,to abstract an operator expression model and devise the workflow engine to complete the visualization construction of the elderly health assessment model,and to detailed design of the main functional modules of the system for the implementation of an elderly health assessment system.This paper will specifically accomplish the following tasks.First of all,aiming at the imperfection of the existing health assessment institutions for the elderly,I am going to analyze the content of the elderly health assessment according to the "Rules for health status assessment in long-term care facilities",ascertain the health assessment indicators for the elderly and calculate the weights of the assessment indicators.I will build the elderly health assessment institutions from the five perspectives of physical health,mental health,living behavioral ability,social adaptability and life habits,in order to achieve a comprehensive assessment to the health status of the elderly.Secondly,in view of the present machine learning methods are not fully applying to the assessment of chronic disease conditions,I studied the performance of different models of machine learning algorithms for building elderly health assessment models on chronic disease data sets.I calculated the importance of feature attributes using the Random Forest method,the correlation between each feature attribute and the target for feature extraction using the Spearman correlation coefficient and evaluated the overall performance of the model using Accuracy,ROC curve,and AUC value.The experimental results show that the Random Forest and SVM methods are better to the construction of elderly health assessment models.Next,because the existing elderly health assessment system cannot meet the needs of medical personnel to dynamically build an elderly health assessment model,I will analyze the workflow of model building and abstract an operator expression model and devise the workflow engine,to build the health assessment model of visualization.By doing so,health care workers will form an elderly health assessment model dynamically and in a quick speed using this system,and assess the health status of the elderly with this well-established health assessment model to provide decision support for medical personnel.Finally,based on the above-mentioned elderly health assessment system and demand analysis of the elderly health assessment system,the system's overall structure,function modules and database will be designed.Through the class diagram,sequence diagram and system interface,there will be a detailed description of the system's function in building an elderly health assessment model and realizing the evaluation to the health of elderly.Also,main functions and performance of the elderly health assessment system can be tested by using the approach of white box and black box testing.