基于模糊理论和欧氏距离自动分析胎儿状态

Automatical Assessment of Fetal Status Based on Fuzzy Theory and Euclidean Distance

胎心宫缩图(CTG)的计算机分析对确定胎儿状态具有重要意义,然而目前基于传统分类标准方法判断的效果不甚理想。为了提高胎儿状态评估准确率,提出了一种新的方法。新方法改进了分类标准,并使用模糊集合来表示CTG参数,从而用得到的集合形成一个特征向量来表示CTG信号,然后计算该信号特征向量与标准状态特征向量之间的欧氏距离,通过比较欧氏距离确定该信号所对应的胎儿状态。实验表明,新方法与专家一结果比较,准确率为88.3%,远高于传统方法的69.9%,而假阳性率仅为7.2%,远低于传统方法的34.9%;与专家二结果比较,准确率为90.3%,远高于传统方法的66.0%,而假阳性率仅为9.0%,远低于传统方法的38.2%。本研究表明新方法有效、可靠。

Computer analysis of cardiotocography （CTG） is very significant to evaluate fetal status. However, current computer analysis based on traditional classification criteria is not ideal. In order to improve the accuracy of fetal status assessment, we proposed a new method. The new method improves the classification criteria and uses fuzzy set to represent the CTG parameters. And then feature vector is formed by that set to represent the CTG signal. By calculating and comparing the Euclidean distance between the signal feature vector and the standard state feature vector, the corresponding fetal status of the signal can be determined. Experiments showed that compared to the results of the first expert, the accuracy rate of new method was 88.3% which was higher than that （69.9%） of the traditional method, and the false positive rate of new method was 7.2% which was much lower than that （34.9%） of traditional methods. While compared to the results of the second expert, the accuracy of new method was 90.3% which was higher than that （66.0%） of the traditional method, and the false positive rate of new method was 9.0% which was well below the 38.2% of the traditional method. Thus the new method is reliable and effective.