Department of Industrial Engineering, Faculty of Engineering, King Abdulaziz University, Jeddah 21589, Saudi Arabia.
Assistant Professor, Department of CSE, RV Institute of Technology and Management, Bengaluru, India.
Comput Intell Neurosci. 2022 Apr 28;2022:9023478. doi: 10.1155/2022/9023478. eCollection 2022.
This study describes a modified approach for the detection of cardiac abnormalities and QRS complexes using machine learning and support vector machine (SVM) classifiers. The suggested technique overtakes prevailing approaches in terms of both sensitivity and specificity, with 0.45 percent detection error rate for cardiac irregularities. Moreover, the vector machine classifiers validated the proposed method's superiority by accurately categorising four ECG beat types: normal, LBBBs, RBBBs, and Paced beat. The technique had 96.67 percent accuracy in MLP-BP and 98.39 percent accuracy in support of vector machine classifiers. The results imply that the SVM classifier can play an important role in the analysis of cardiac abnormalities. Furthermore, the SVM classifier also categorises ECG beats using DWT characteristics collected from ECG signals.
本研究描述了一种使用机器学习和支持向量机（SVM）分类器检测心脏异常和 QRS 复合体的改进方法。该方法在敏感性和特异性方面均优于现有方法，对心脏不规则的检测错误率为 0.45%。此外，向量机分类器通过准确地对四种心电图（ECG）拍类型进行分类，验证了该方法的优越性：正常、左束支传导阻滞（LBBB）、右束支传导阻滞（RBBB）和起搏拍。该技术在多层感知机（MLP）-BP 中的准确率为 96.67%，在支持向量机分类器中的准确率为 98.39%。结果表明，支持向量机分类器可以在心脏异常分析中发挥重要作用。此外，支持向量机分类器还可以使用从 ECG 信号中采集的 DWT 特征对 ECG 拍进行分类。
