Abstract: Early detection of hypertension is essential as it is a common chronic age-related disease often associated with debilitating cardiovascular complications. The aim of this article is to find the best accuracy performance in detecting hypertension. This article introduced a method for hypertension detection without a cuff using parameters of photoplethysmography (PPG) and support vector machine (SVM). The parameters were heart rate variability (HRV) of PPG. Both the time domain and frequency domain of HRV were utilized. The HRV was obtained from the respiratory rate (RR) interval of PPG, which was the time interval between two consecutive peaks of PPG. An SVM with a radial basis function (RBF) was used. SVM parameters were tuned to find the optimal one. Furthermore, a feature selection of PPG-HRV was conducted to find appropriate features. Experiments using clinical data showed that SVM with HRV of PPG resulted in a good performance for hypertension detection. The performance of hypertension detection with different values of SVM parameters and different features of HRV was presented. This method found accuracies of 98.89 and 80.68% for training and testing, respectively. Based on the results, the use of HRV from PPG is quite effective and can contribute to medical development to detect hypertension.
Citation: Octaviani, A., Nuryani, N., Salamah, U., Utomo, T.P. (2023). Heart Rate Variability of Photoplethysmography for Hypertension Detection Using Support Vector Machine. In: Triwiyanto, T., Rizal, A., Caesarendra, W. (eds) Proceeding of the 3rd International Conference on Electronics, Biomedical Engineering, and Health Informatics. Lecture Notes in Electrical Engineering, vol 1008. Springer, Singapore.
DOI: https://doi.org/10.1007/978-981-99-0248-4_31