Construction And Demo Applications Of TCM Symptoms Ontology

With the new information technologies gradually infiltrated the hospital,such as Internet of Things,cloud computing,artificial intelligence,mobile Internet,the hospital information system collect and store massive medical data about the health services.Those truly reflect the medical practice activities.The concept of big data is proposed,which makes people aware of association analysis on the hospital's massive data is more valuable than a small sample of data analysis.Therefore,using real-world massive data to carry out Chinese medicine-related studies,speaking and deciding with data has become a new research trend.Symptoms are core data of TCM syndrome differentiation and treatment.It is TCM doctor using natural language to record medical information during recognizing disease and diagnosing so that the expression of symptoms is difficult to unify.That causes sharing and depth using of symptoms by computer technical difficult to achieve.Therefore,building a computer-readable,operational and normalized,formalized conceptual model is necessary and urgent.Objective: Building TCM Symptoms Ontology(except symptoms of tongue diagnosis and Pulse-taking)as a foundation of understanding and processing symptoms by the computer.Exploring new ways to process and use symptoms.Methods: This research adopts the seven_steps method to construct symptoms ontology seven-step method.Based on literature research,getting and combing terms of symptoms.Building the symptoms ontology model by the mix of "Bottom-up" and "top-down" method,that includes dividing ontology class,getting its attributes and constraints.Then formalized express symptoms ontology with Classification and coding technology,and building instances base of symptoms ontology.Finally,we develop symptoms data acquisition system.With help of it,we get symptoms data of 114 jaundice medical records and analys the data to verify the feasibility of the model.Results:(1)Having constructed symptoms ontology model,which have ontology class based on TCM diagnostic method symptoms and 23 class properties that are body concept,nature,extent,occurrence factors,aggravating factors,mitigating factors,the onset of urgency,seizure frequency,duration,population,location,smell,period,location,texture,color,color,shape,dynamics,volume,times,discharge flu,disease,etc.Those can be divided into common attributes and specific attributes.The relationships of the symptoms are synonymous relationship,superior and lower relationship;(2)Having constructed instance base of symptoms ontology.Taking ontology classes and attributes as 24 category axis of symptoms by using "ICD" multiaxial classification thought.And using classification and coding techniques to code symptoms instances.(3)Proposed that there is a hierarchy in axis of location,color,syndromes,occur factors,aggravating factors,mitigating factors,and approach of symptoms.With reference to the method of dividing ontology class bulid the corresponding level hierarchy;(4)Developped the symptoms data acquisition system in C/S mode.Expored a application path of symptoms ontology,which is recording symptom information in human-computer interaction mode,saving the data in a structured and standardized form,and using data to study in multi-angle data mining.Verify the feasibility and openness of symptoms ontology model.Conclusions: Symptoms ontology expressed symptoms connotation in a standardized and structured way,which provides a basis for computer understands and semantic processing the symptoms.At the same time,developed the symptoms data acquisition system and used quickly enter information technology to solve the contradiction between full and long time of symptom information collection.In addition,we find a effective way to explore the depth of the value of symptoms.However,the symptoms ontology model in this study did not take the symptoms of tongue diagnosis and Pulse-taking into account,and the presence of property classification is not perfect ethier,negative symptoms and muti-symptoms can not be expressed.These proplems will be explored in the next research.