Research And Application Of Orthodontic Case Similarity Analysis Method

The emergence of electronic medical records has improved the efficiency of inductive management of case records(cases).At the same time,studying the similarity of cases has important guiding significance for doctors in the treatment process.It is difficult to extract the information from medical records as characteristic factors.This paper studies the similarity measurement methods of orthodontic cases.To study the similarity of orthodontic cases,we need to analyze the information recorded in the cases and determine the characteristic descriptors of the cases.At present,when analyzing the similarity of cases,the diagnosis information and the treatment plan are used as the characteristic descriptors of the case.Considering the inaccuracy of the diagnosis information and the treatment plan in the orthodontic cases,this paper select measurement data is taken as characteristic descriptors,it details the measured values of various parameters.Experiments show that the characteristic descriptors can retrieve similar cases more accurately.When measuring the similarity of the cases,considering the excessive measurement parameters,the complexity of the algorithm can be increased if they are used to calculate the similarity coefficient.Therefore,the measurement information is divided into three parts according to the correct standard parameters,the single tooth parameters and the whole jaw parameters,to study cases similarity.In the calculation process,the distance similarity matching algorithm requires the data units to be consistent,and the parameters in the measurement information do not meet this condition.Aiming at this problem,this paper classifies the measured information data units into one category,calculates its similarity distance and takes the mean value as the distance similarity coefficient of the final case.In addition,a ratio algorithm is presented,which maps different parameters into length or angle,and calculates similarity coefficient by occupying ratio.Experiments show that the ratio algorithm calculated by the proposed algorithm is basically consistent with the distance similarity algorithm.In practical applications,this paper selects three different similarity algorithms for similar case retrieval,each algorithm gives the most similar first three cases in the case library,and the corresponding 3D dental models,to improve the accuracy of the retrieval and the friendliness of the interface.In the study of orthodontic cases similarity,the similarity algorithm was realized by MyEcilpse and Java programming,the 3D model display was realized by JavaScript and the Three,js library.Through practical application,the method presented in this paper can retrieve similar cases quickly and accurately.