@misc{oai:mie-u.repo.nii.ac.jp:00015127,
 author = {Félix, Chávez José Andrés},
 month = {Mar},
 note = {application/pdf, Cardiovascular diseases are one of the main causes of mortality around the world. Many of these events happen outside-of-hospital conditions when the person is at home or at a public space without being monitored. Advances in technology have improved the patients’ monitoring, be it a Holter monitor or a wearable device. However, these devices are still sensitive to noise from movement, therefore it is important to understand how the movement affects electrocardiographic recordings.
In this research three devices were tested. The BIOPAC system is a nonmobile device that represented conventional electrocardiography. The Hitoe system and the Vitalgram system are wearable solutions for ECG monitoring.
Since the goal is to understand how noise from motion affects the ECG signal, based on previous research the motion noise was determined to be the sum of baseline wander (BW), electromyogram (EMG), and electrode motion artifacts (EMA). BW and EMG were obtained by using a low-pass filter and a high-pass filter, respectively. The EMA were obtained by doing a beat-by-beat analysis. To analyze the signal beat-by-beat, the R- peaks of the ECG signal had to be identified. To do this, the discrete wavelet transform (DWT) was employed.
Once all the types were extracted, the signal-to-noise ratio (SNR) was used as a parameter to determine the quality of the ECG recordings from each device. The subjects were shown a video and were asked to perform certain actions in resting and stress conditions. The SNR values were more negative when the subject performed physical activities.
The results show that the ECG recordings from these devices were not affected by EMG noise, since the SNR values for EMG were always above 30 dB. As for the BW noise, the Vitalgram device was not affected, but the Hitoe and BIOPAC systems were mildly affected by it. The Vitalgram system was the only device to present a positive average SNR value for EMA while the subjects were running.
In Apandi’s research the performance of several heartbeat detection algorithms was tested, including Apandi’s proposed method. Using Apandi’s research data and the SNR values from this research, the performance of these algorithms on recordings from these devices was estimated. The results show that using Apandi’s proposed algorithm on recordings from the Vitalgram device would render the best results, since the estimated heartbeat detection performance was of 100%., Human Support Systems Laboratory Department of Mechanical Engineering Graduate School of Engineering Mie University, 53p},
 title = {Analysis of the Noise Characteristics of Ambulatory Electrocardiogram},
 year = {2022}
}