摘要
针对长期积累的病历数据不仅数量庞大,记录方式、内容千差万别,而且噪声、缺省值大量存在的问题,提出了智能医疗诊断系统,利用以往的病历数据经过数据挖掘等技术来产生决策规则,以期为潜在的患者及早发现病情,获得早期诊治。总结了病历文本的记录内容和特点,对每一类的特点和形成原因进行了分析研究,并结合医疗诊断分别提出了针对它们的预处理方法,为下一步数据挖掘做好了准备。
Intelligent healthcare system data preprocess problem is discussed.It is to help prospective patient find his condition as soon as possible with decision rules produced based on medical cases by using data mining technology. Medical cases are not only huge in amount, but also different from each other in record styles and contents.The content and characteristic of medical cases is summarized.Combining with medical diagnosis, it aims at each kind to bring forward the idea of data preprocess for data mining respectively. The proposed method makes good preprocess for data mining.
出处
《控制工程》
CSCD
2005年第1期33-35,66,共4页
Control Engineering of China
基金
澳门大学科研委员会资助项目(RG049 02 03S VMI FST)
