基于PCA的Bayes分类器应用于心电图临床诊断
Naive Bayes Classifier Based on Principal Component Analysis Applied to Clinical Diagnosis Decision of ECG Data

心血管疾病(CVD)是一种常见的慢性疾病，初期没有明显症状，发展较慢难以发现，且发病危险性高。因此检查环节十分重要，其中动态心电图采集大量心电数据，大量的心电数据在支持各种心脏疾病诊断的同时，却也提升了人力物力的分析成本，需耗费大量医师人力，大大降低诊断效率，加上人工检查可能因疲劳或分心产生失误，降低可靠性。为了更有效率的运用医师人力、减少人工误差、提高医疗水平质量、解放医师人力与更有效的运用医疗资源以惠及广大患者，设计出了如何利用患者的ECG数据：对资料进行特定的预处理，接着将数据汇入Matlab和SPSS进行主成分分析，之后使用贝叶斯分类器对机器进行训练，并给出运行诊断的结果准确率为75.8%。
Cardiovascular disease (CVD) is a common chronic disease with no obvious symptoms at the initial stage, slow development and difficult to detect, and high risk of morbidity. The check process is very important, dynamic electrocardiogram (ecg) collects ecg data, large amounts of ecg data are used in support of various heart disease diagnosis, which promotes the analysis of the manpower cost, costs a lot of physician manpower, greatly reduces the efficiency of diagnosis, and artificial check may cause failure or lower the reliabilitydue to fatigue or distraction. In order to use the physician manpower more efficiently, to reduce the manual error, to improve the quality of medical treatment, to liberate the physician manpower and to use medical resources more effectively to benefit the patients, how to use the ECG data of patients is designed: after a specific preprocessing of the data, the data was imported into Matlab and SPSS for principal component analysis. After that, the machine was trained by Bayesian classifier and the accuracy rate of operation diagnosis was 75.8%.