Loading...

Design and Implementation of Non-Invasive Blood Pressure Monitor for Smart Health Application

Amini, Alireza | 2020

363 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 52794 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Fakharzadeh Jahromi, Mohammad
  7. Abstract:
  8. The blood pressure (BP) is one of the most vital signs to monitor human health. Hypertension is a major public health problem that is associated with morbidity and mortality. In the past decades, the number of people with hypertension disease has been increased. According to Ministry of Health, 54 percent of hypertensive patients are unaware of this health threat. Therefore, accurate and continuous measurement of BP is essential for its prevention and treatment. Traditional non-invasive methods of BP monitoring use an inflatable cuff, which causes discomfort due to occlusion of the artery and can not be used continuously. Numerous studies have attempted to present methods without using a cuff. Most of them involved the use of a parameter called pulse transient time, which is mainly calculated from electrocardiogram (ECG) and photoplethysmogram (PPG). Such methods require periodic calibration. In this study, in order to address the calibration issue, by relying on deep learning and convolutional neural networks, BP related features extracted from ECG and PPG signals and finaly blood pressure is calculated. By using an online database, the proposed method is evaluated by AAMI, BHS and IEEE 1708 standards. From the AAMI standard point of view, the proposed method falls within the acceptable range for estimating diastolic blood pressure. Compared to BHS standard criteria, the proposed method receives grade A for DBP measurement and grade B for SBP. Also, compared to IEEE 1708, the accuracy of this method for DBP is in the grade A range and for SBP is in the range of grade C. The proposed method was evaluated by collecting samples from different individuals using BIOSEN wristband
  9. Keywords:
  10. Electrocardiogram ; Deep Learning ; Convolutional Neural Network ; Photoplethysmography ; Blood Pressure Measurement ; Continuous Measuremnt ; Non-Invasive Blood Pressure Measurement

 Digital Object List

 Bookmark

No TOC