计算机明尼苏达编码与人工编码比较及其评价  

Comparison of the performance of computerized Minnesota coding and manual coding.

作　　者：吴杰[1] J.A Kors G.Herpen Z.Zhang R.J Prineas 

机构地区：[1]同济医科大学附属同济医院心内科,430030 [2]荷兰Erasmus大学医学信息研究所 [3]美国Miami大学公共卫生和流行病学研究所

出　　处：《中华心律失常学杂志》1998年第2期92-95,共4页Chinese Journal of Cardiac Arrhythmias

摘　　要：目的介绍一种计算机明尼苏达编码系统,并报告计算机编码与人工编码性能的比较结果。方法随机选择300例有各种编码形态的心电图,分别独立由计算机和人工进行编码并与参考标准编码比较,计算两者编码的准确率以及检测不同编码分类的敏感性和特异性。结果研究结果显示在所有的编码分类上计算机编码的准确率以及敏感性和特异性优于或相当于人工编码。结论本研究表明计算机编码是一项有价值的编码方式,可以辅助或替代人工编码。Objective The Minnesota code (MC) is an electrocardiogram(ECG) classification sys- tem for epidemiologic study.As the MC measurement procedures and rules are complex,manual coding is time consuming and error prone.The authors developed a computerized MC system,closely adhering to the MC regulation.This study compared the performance of the computerized coding and manual coding.Meth- ods A test set of 300 ECGs containing a wide variety of codable patterns was collected.The ECGs were coded independently by the computer and by an experienced human reader.A reference code'truth'was es- tablished by resolving disagreements through a consensus procedure.If the computer and human agreed, they were considered to be correct.The total agreement was compared with reference code,both for the computer and for manual coding.Sensitivity and specificity were also computed for each of nine main code categories of the MC.Results The results shows that computer coding is as good as or better than the hu- man reader for the accuracy,sensitivity and specificity of all the MC categories.Conclusion The comput- erized Minnesota coding is a valuable alternative or supplement to manual coding.

关 键 词：人工 心电图 计算机编码 特异性 敏感性 辅助 研究结果 参考标准 准确率 评价 

分 类 号：R2-03[医药卫生—中医学] R541