MUHAMMAD KASHIF (2022) SCALOGRAM DRIVEN HYPERTENSION CLASSIFICATION USING PHOTOPLETHYSMOGRAPHY (PPG) SIGNAL: A DEEP LEARNING APPROACH. Thesis thesis, Universitas Airlangga.

	Text (HALAMAN JUDUL) 082015053001_HALAMAN JUDUL.pdf Download (252kB)
	Text (BAB I) 082015053001_BAB I.pdf Download (214kB)
	Text (FULLTEXT) 082015053001_FULLTEXT.pdf Restricted to Registered users only Download (5MB) | Request a copy

Abstract

Blood pressure is a basic physiological parameter in the cardiovascular circulatory system. Long-term abnormal blood pressure will lead to various cardiovascular diseases, making the early detection and assessment of hypertension profoundly significant for the prevention and treatment of cardiovascular diseases. In this study, we investigate whether or not deep learning can provide better results for hypertension assessment when compared to the classical signal processing and feature extraction methods. We tested a deep learning method for the classification and evaluation of hypertension using photoplethysmography (PPG) signals based on the continuous wavelet transform (using Morse) and convolutional neural network. We collected 163 data recording, with 74 healthy subjects (38 male and 36 female) and 89 subjects were detected as hypertension people (20 male and 69 female). According to the seventh report of the Joint National Committee, blood pressure levels are categorized as normotension (NT), prehypertension (PHT), and hypertension (HT). In order to assess the hypertensin by using PPG signals, twelve classification experiments were conducted. The classification experiments were including: (1) Normotension vs Prehypertension; (2) Normotension vs Stage-1 Hypertension; (3) Prehypertension vs Stage-1 Hypertension; (4) Normotension vs Prehypertension vs Stage-1 Hypertension; (5) Classification based on Pulse Pressure; (6) Classification based on systolic blood pressure; (7) Classification based on diastolic blood pressure; (8) Classification based on brachial systolic blood pressure; (9) Classification based on brachial diastolic blood pressure; (10) Classification based on aortic systolic blood pressure; (11) Classification based on aortic diastolic blood pressure; (12) Classification based on Pulse Wave Velocity. The validation accuracy of these twelve classification trials were 91%, 92%, 97%, 91%, 92%, 82%, 88%, 84%, 86%, 88%, 89% and 93%, respectively. The tested deep method achieved higher accuracy for hypertension assessment when compared to the classical signal processing and feature extraction method. Additionally, the method achieved comparable results to another approach that requires electrocardiogram and PPG signals.

Actions (login required)

View Item