Skip to Main Content
 

Global Search Box

 
 
 
 

Files

File List


Thesis_Final_CerenUyanikCivek.pdf (2.84 MB)

ETD Abstract Container

Abstract Header

Signal Quality Assessment of Photoplethysmogram for Heart Rate Estimation

Uyanik Civek, Ceren
http://rave.ohiolink.edu/etdc/view?acc_num=osu1593651500804312

Abstract Details

2020, Master of Science, Ohio State University, Electrical and Computer Engineering.
Wearable devices make a significant contribution in health monitoring systems by continuously collecting important health parameters with the intention of prevention, early detection and control of diseases. Photoplethysmography (PPG) appears as a feasible, affordable and non-invasive technique for long term health monitoring and is commonly used in various wearable systems and medical devices. However, PPG signal is highly prone to different noise sources such as motion and ambient light which affect the reliability of the estimated results of health metrics. Therefore, a measure of quality, signal quality index (SQI), is needed for identifying poor signals distorted by artifacts before making any interpretations. In this study, three SQI estimation methods are presented in order to assess PPG signals collected from MotionSense HRV. First method, adaptive template matching, generates individual-specific templates and calculates correlation for signal quality. Second method, subspace matching, focuses on the linear combination of high quality PPG signals, while autoencoder as a third method provides the nonlinear correspondences between them. The performances of proposed approaches are evaluated based on heart rate measurements obtained from both PPG and ECG signals with the hypothesis that the error in the heart rate estimation reduces as the poor data is discarded. In addition, as an application of SQI, the analysis of sensor design in terms of channels and light color is performed. All three methods validate the hypothesis by showing that the error monotonically decreases during data rejection. The results also indicate that the choice of light and the combination of channels affect measured data quality. Combining channels significantly reduces the error with less data losses rather than using individual channels. Additionally, green color is more effective than infrared color in assessing signal quality.
Emre Ertin (Advisor)
Lee Potter (Committee Member)
81 p.
Electrical Engineering
Photoplethysmogram; Signal Quality Index; Wearable Sensors; Heart Rate Estimation; Multi-modal Signal Quality Assessment

Recommended Citations

Citations

  • Uyanik Civek, C. (2020). Signal Quality Assessment of Photoplethysmogram for Heart Rate Estimation [Master's thesis, Ohio State University]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=osu1593651500804312

    APA Style (7th edition)

  • Uyanik Civek, Ceren. Signal Quality Assessment of Photoplethysmogram for Heart Rate Estimation. 2020. Ohio State University, Master's thesis. OhioLINK Electronic Theses and Dissertations Center, http://rave.ohiolink.edu/etdc/view?acc_num=osu1593651500804312.

    MLA Style (8th edition)

  • Uyanik Civek, Ceren. "Signal Quality Assessment of Photoplethysmogram for Heart Rate Estimation." Master's thesis, Ohio State University, 2020. http://rave.ohiolink.edu/etdc/view?acc_num=osu1593651500804312

    Chicago Manual of Style (17th edition)

osu1593651500804312
617
© 2020, all rights reserved.
This open access ETD is published by The Ohio State University and OhioLINK.
Release 3.2.12