A sudden cardiac death predictor using the spatial QRS-T angle feature combined with QRS duration and a support vector machine achieved 100% accuracy in classifying healthy controls and cardiac patients.
Case-Control (n=18)
Does a predictor based on spatial QRS-T angle and SVM accurately classify patients with cardiac disease versus healthy controls?
A machine learning model using spatial QRS-T angle and QRS duration features achieved 100% accuracy in a small sample for detecting cardiac disease.
This paper presents a case study of sudden cardiac death predictor based on spatial QRS-T angle (spQRSTa) feature and support vector machine (SVM). The comparison between common ECG features and spQRSTa feature is presented in the paper. Eighteen volunteers were involved in the ECG experiement. The ECG data consist of 8 healthy controls and 8 patients with cardiac disease. Four ECG features such as QRS duration, QT interval, QT correction and spQRSTa feature were extracted from raw ECG signal. The two pair combination of 4 features were presented. The results show that the pair combination where the spQRSTa feature included can distinguish the healthy controls and patients with cardiac disease. From the plot results, the pair between QRS feature and spQRSTa feature was selected for SVM classification. The result show that SVM can classify the two classes with 100% accuracy.
Caesarendra et al. (Thu,) conducted a case-control in Cardiac disease (n=18). Spatial QRS-T angle (spQRSTa) feature and support vector machine (SVM) vs. Common ECG features (QRS duration, QT interval, QT correction) was evaluated on Classification accuracy between healthy controls and patients with cardiac disease. A sudden cardiac death predictor using the spatial QRS-T angle feature combined with QRS duration and a support vector machine achieved 100% accuracy in classifying healthy controls and cardiac patients.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: