Subject: Photoplethysmogram data · Signal processing · Heart rate analysis · Peak detection · Evaluation metrics
Year: 2021
Type: Proceedings
Title: Evaluation of Python HeartPy Tooklit for Heart Rate extraction from PPG
Author: Hristina, Mitrova
Author: Koteska, Bojana
Author: Madevska Bogdanova, Ana
Author: Lehocki, Fedor
Author: Ondrusova, Beata
Author: Ackovska, Nevena
Abstract: Handling the mass casualty emergency situations can be improved by introducing a chest patch sensor that is able to deliver the main vital parameters: Heart Rate (HR), Respiration Rate (RR), SPO2 and Blood Pressure. The START triage procedure requires both HR and RR parameters almost instantly. In this paper we investigate the calculation of HR from a raw PPG signal, using appropriate functions from the Python HeartPy Tooklit, by comparing the calculated HR to the measured HR for the same patients, recorded at the same time as the PPG signal. By using several evaluation metrics, it was concluded that there is no significant difference between the measured and the calculated HR (MAE = 0,3, MSE=0,3, R2 =0,99, Pearson’s and the Spearman’s coefficient of correlation, 0.99). This result is the same whether raw or filtered PPG signal was used for the HR calculation.
Publisher:
Relation: ICT Innovations Conference 2021
Identifier: oai:repository.ukim.mk:20.500.12188/22821
Identifier: http://hdl.handle.net/20.500.12188/22821