SPHERE wrist ECG data

Published: 24 September 2018| Version 1 | DOI: 10.17632/tzfmydyzch.1
Alex Casson, Christopher Beach


Description of Data: - ECG data as recorded from the wrists for the IEEE Access paper ‘An ultra low power personalizable wrist worn ECG monitor integrated with IoT infrastructure’. - Data is recorded for 5 participants, ages 23 – 36. - Sampling rate is 128 Hz. Files are saved in the Matlab .mat format.


Steps to reproduce

Steps to reproduce: Using the custom ECG amplifier and electrodes described in IEEE Access paper, data is sampled from 5 participants sitting still at a desk. Data has been processed in Matlab, using the first 4 signal processing steps described in the paper in section V, subsection B. These are: - Low pass (Fl = 30 Hz), high pass (Fh = 0.1 Hz), and notch filtered (Fd = 47.5 Hz, Fu = 52.5 Hz), using first order zero phase delay (filtfilt in MATLAB) Butterworth filters). - Remove ECG baseline wonder using Discrete Wave Transform. - Smooth ECG using extended Kalman filter.


The University of Manchester, University of Bristol


Health Sciences, Biomedical Engineering