Loading...

Non-invasive and Continuous Blood Pressure Measurement Using Pulse Transit Time without Cuff Requirement

Esmaili Dastjerdi, Amir Hossein | 2017

809 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 49713 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Shabany, Mahdi
  7. Abstract:
  8. This thesis addresses cuff-less blood pressure (BP) estimation, for systolic blood pressure (SBP) and diastolic blood pressure (DBP) estimation with the aid of the processing of vital signals. Both with and without calibration per person approaches for BP estimation are considered. A data acquisition hardware is designed for high-resolution sampling of phonocardiogram (PCG), Photoplethysmogram (PPG) and electrocardiogram (ECG). Using these signals, pulse transit time (PTT) and pulse arrival time (PAT) indexes, which would be used for BP estimation with calibration, are extracted. For the BP estimator model using the calibration, A nonlinear model based on PTT or PAT, via modeling the vessels as elastic tubes, is introduced. To derive the parameters of this model, subjects undergo a supervised physical exercise which causes their BP rise and then, a number of reference BPs are measured alongside the acquisition of the signals per subject. It is suggested to use a force sensing resistor (FSR) that is placed under the cuff of the BP reference device to mark the exact moments of reference BP measurements during the acquisition of signals, which are corresponding to the inflation of the cuff. The proposed calibration-based method is evaluated for 32 subjects, where there are 5 or 6 reference BP measurements for each subject. For BP estimation without calibration, From 69 subjects, PPG signals and reference BP measurements are acquired. Using the features extracted from the PPG signal, and by exploiting maching learning models, BP estimation without calibration is performed. By evaluating the proposed methods for BP estimation, and comparing them with other works on non-invasive BP estimation, the results are promising. In the estimation approach with calibration, using PTT index, mean absolute errors for SBP and DBP estimation are 6.90 mmHg and 4.15 mmHg, respectively. Using PAT index, these values are 5.07 mmHg and 4.52 mmHg, respectively. Mean absolute errors for SBP and DBP estimation, without calibration per person, are 9.41 mmHg and 6.78 mmHg, respectively, which shows in general, estimations with calibration, with the cost of performing calibration procedure per person, can provide better accuracies. The results show that the proposed methods, provide a reliable estimation of SBP and DBP, and can be utilized in mobile healthcare systems
  9. Keywords:
  10. Cuffless Blood Pressure Estimation ; Pulse Arrival Time (PAT) ; Pulse Transit Time (PTT) ; Blood Pressure ; Mobile Healthcare System ; Biological Signals

 Digital Object List

 Bookmark

No TOC