基于本体的医疗自动诊断系统的研究与应用

发布时间：2018-07-07 17:03

本文选题：本体 + 自动诊断　；参考：《电子科技大学》2017年硕士论文

【摘要】：随着我国人口数量地增多,人口老龄化趋势越来越严峻,同时人们生活水平不断地提高,人们更关注自己的生命健康状况。目前,医生依旧采取问诊和医学检查相结合的方式对病人进行疾病诊断,但是,我国医疗水平发展在不同地区、不同医院之间存在不均衡的问题,医生的能力也存在差异,许多病人为了确诊所患的疾病将花费高昂的费用。如何根据症状来自动化地诊断病人所患疾病是本文研究的重点。在过去的二十多年,本体已经广泛应用到知识工程、人工智能、信息推荐、自然语言处理、生物信息学、农业领域等工程。基于本体的应用越来越多,然而,许多情况下本体是手工构建,存在工作量大、效率低、关系表达错误等缺点。本文在形式概念和本体理论的基础上,把本体作为知识表达和共享的载体,将症状、疾病知识组织起来,建立疾病与症状的本体知识库,以便实现医疗的自动诊断。综上所述,本文研究的主要内容如下:1.在构建疾病与症状的本体过程中,改进了属性偏序结构图算法,并利用该算法自动化地构建疾病本体知识库。针对疾病数据集,提出了症状信息量的概念来刻画具有某些症状的疾病数量,并选取具有最小信息量的公共症状集合,构建属性偏序结构图。实验结果表明,利用最小信息量的公共症状集合构建偏序结构图,降低了多个症状之间的冗余性。2.针对疾病数据集,分析了构建的疾病与症状的本体结构,提出了一套用于病人疾病诊断的医疗诊断模型。在医疗诊断模型中,利用症状之间的相似度与症状的权重进行加权平均,计算出病人疾病与该疾病之间的相似度,并利用该相似度来衡量病人患有该疾病的概率。实验结果表明,该算法在急性膀胱炎和急性肾炎的疾病数据中进行医疗诊断的准确率都在80%以上。3.设计并实现了医疗自动诊断系统。该系统使用MVC设计模式,使系统界面显示和疾病诊断过程相互独立。在进行医疗自动诊断时候,医生采用问诊的形式,确定病人具有的症状,系统自动根据病人的症状进行医疗诊断并输出诊断结果。经测试,该系统能够便捷地对病人进行医疗诊断。
[Abstract]:With the increase of population in our country, the aging trend of population is becoming more and more serious, and people's living standard is improving constantly, so people pay more attention to their own life and health. At present, doctors still use the combination of examination and medical examination to diagnose patients' diseases. However, the medical level of our country is developing in different regions, there are uneven problems among different hospitals, and the ability of doctors is also different. Many patients will be expensive to diagnose the disease. How to diagnose the patient's disease automatically according to the symptoms is the focus of this paper. In the past twenty years, ontology has been widely used in knowledge engineering, artificial intelligence, information recommendation, natural language processing, bioinformatics, agriculture and so on. There are more and more applications based on ontology, however, in many cases ontology is constructed by hand, which has the disadvantages of heavy workload, low efficiency, wrong expression of relationship and so on. Based on the formal concept and ontology theory, this paper takes ontology as the carrier of knowledge expression and sharing, organizes the knowledge of symptoms and diseases, and establishes the ontology knowledge base of diseases and symptoms in order to realize the automatic diagnosis of medical treatment. To sum up, the main contents of this study are as follows: 1. In the process of constructing the ontology of disease and symptom, the algorithm of attribute partial order structure graph is improved, and the knowledge base of disease ontology is constructed automatically by using this algorithm. According to the disease data set, the concept of symptom information quantity is proposed to describe the number of diseases with some symptoms, and the common symptom set with minimum information is selected to construct the attribute partial order structure. The experimental results show that the least amount of common symptom set is used to construct the partial order structure chart, which reduces the redundancy of multiple symptoms. 2. According to the disease data set, this paper analyzes the ontology structure of the disease and symptom, and puts forward a set of medical diagnosis model for the patient's disease diagnosis. In the medical diagnosis model, the similarity between the symptoms and the weight of the symptoms is weighted average, and the similarity between the disease and the disease is calculated, and the probability of the patient suffering from the disease is measured by the similarity. The experimental results show that the accuracy of the proposed algorithm in the diagnosis of acute cystitis and acute glomerulonephritis is over 80%. A medical automatic diagnosis system is designed and implemented. The system uses MVC design pattern to make the system interface display and disease diagnosis process independent of each other. In automatic medical diagnosis, doctors use the form of inquiry to determine the symptoms of patients, the system automatically according to the symptoms of patients for medical diagnosis and output diagnosis results. The test results show that the system can be used to diagnose patients conveniently.
【学位授予单位】：电子科技大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：R318;TP391.1

【参考文献】