【作者】 庄军；

【导师】 郭平；

【作者基本信息】 重庆大学 ， 计算机应用技术， 2006， 硕士

【摘要】 近年来,随着医疗卫生事业的发展,在临床医疗活动中的药物使用问题逐步成为公众关注的焦点。据统计资料显示,中国每年五千多万住院人次中与药物不良反应有关的可达二百五十多万人,死于药物不良反应的约有近二十万人,药物致死占死因第4位,其危害不容低估。随着药物的广泛使用,药源性疾病对人类健康的危害也越来越大,合理使用药物是降低药物不良反应危害的重要前提。合理用药是指根据疾病种类、病人状况和药理学理论选择最佳的药物及其制剂,制定或调整给药方案,以期有效、安全、经济地防治和治愈疾病的措施。解决临床合理用药中存在的技术性问题的指导性原则就是循证医学。循证医学的主要研究方法有随机对照研究、系统评估和临床指引等几种。本文通过对数据挖掘中关联规则挖掘技术、数据预处理技术和专家系统的深入研究,根据我国医疗卫生部门信息化发展的现状和趋势,结合临床医疗活动的特点,提出一种基于临床合理用药的药物治疗专家系统模型,并对实现进行了探索。该系统通过对医院信息系统下电子病例资料的数据抽取和医嘱信息的提取,在进行必要的数据预处理后按年龄、性别等分类,形成相应的医嘱序列,再运用改进的SCG算法挖掘,可以得到一组或多组医嘱序列,这种医嘱序列和案例序列一起构成临床药物治疗知识库,为临床医生在给患者治疗过程中药物治疗方案的制定提供决策支持。本文通过对Apriori算法和SCG算法的研究,提出SCG算法的改进方法,实验证明,改进后的SCG算法比SCG具有更高的挖掘时间效率。本论文首次将数据挖掘技术应用于临床合理用药的循证研究,采用数据挖掘技术提高了临床合理用药知识的自动收集和验证能力,促进临床医学领域数据挖掘技术的应用发展,有效地提高临床合理用药指导和监督的智能化水平。更多还原

【Abstract】 In recent years,drug threpty in becomes the focus of public attention with the development of medical and healthy services. According to the statistical data,there are more than two million and five hundred thousand involved in drug-induced disorders in fivty million in hospital, and two hundred thousand persons died of badness drug threpty. Drug-induced disorder becomes the Number four reason causing death, so it admits of no underestimate. As the widely used drug, drug-induced disorder is a great hazard to health of human. Rational administration of drug is to reduce hazard of bad effect of the medication.Rational administration of drug means making best choice of medicine and the preparation according to sickness category, conditions and pharmacology theory in order to get the valid, secure and economic measure to prevent and charm away disease.The guide principles to resolve techno-problems in clinic rational administration of drug is Evidence-based Medicine, in which the main research methods are random contradistinguish research, system evaluation, clinic direction and so on.Based on the deep research of association rules of data mining, data pretreatment and, actuality and trend of sanitation department in our country, the thesis proposed an expert system model and explored its realization combing with characteristic of medical treatment activity in our country. The system get a set of doctor’s advice by picking-up the data of CPR and doctor’s advice, sorting by age or sex after necessary data pretreatment, formint the corresponding doctor’s advice lists, using the improved SCG algorithm. The doctor’s advice and case lists compose clinic medical cure repository, which support decision-making for clinic cure scheme.According to the algorithm of Apriori and SCG, in this paper, we propose the improved SCG algorithm, the experiment proment that the improved SCG algorithm is more efficitive than SCG algorithm.This paper applies data mining in clinic rational administration of drug for the first time and advance power of collection automatically and validation in clinic rational administration of drug. It will boost development of application of data mining in clinic astrology field and advance level of reasonable drug threpty and intelligentization.更多还原