基于PIX和RIM的W-EMPI算法研究及其在医疗中的应用

Research and Application of W-EMPI Algorithm Based on PIX and RIM in Medical

历史患者与新入院患者的身份匹配是医疗电子病历内容共享领域面临的首要问题。针对现有算法存在的流程简单、缺乏优化等缺点,提出一种基于PIX技术和RIM模型的W-EMPI算法。首先通过数据清洗得到标准数据集;然后利用唯一特征信息筛选去除实名数据;最后通过相似度评价的方式进行患者身份匹配,实现对相似病人的合并定位。算法在江苏省某大型三甲医院的EMPI系统中进行了初步应用,实践效果表明,与传统方法相比,所提算法具有更高的准确率,能够更加有效地完成患者匹配。

Identity matching between historical patients and newly admitted patients is the primary problem in the field of medical electronic medical record content sharing.Aiming at the shortcomings of existing algorithms such as simple process and lack of optimization,a W-EMPI algorithm based on PIX technology and RIM model is proposed.Firstly,the standard data set is obtained by data cleaning.Then the unique feature information is used to filter out the real-name data.Finally,patient identity matching is carried out through similarity evaluation to achieve the combined positioning of similar patients.The algorithm has been preliminarily applied in the EMPI system of a large tertiary hospital in Jiangsu Province.The practical effect show that compared with the traditional procedure,the proposed algorithm has higher accuracy and can more effectively complete patient matching.