Measurement and analysis of physiological parameters using signal processing techniques
Health is very essential in everyone's life but to always stay healthy, it becomes a very challenging task especially for the citizens of developing countries. To have a good health, it is important to monitor the physiological parameters such as heart beat rate/pulse rate, blood pressure, blo...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Language: | English |
| Published: |
2018
|
| Subjects: | |
| Online Access: | https://eprints.ums.edu.my/id/eprint/22635/1/Measurement%20and%20analysis%20of%20physiological%20parameters%20using%20signal%20processing%20techniques.pdf https://eprints.ums.edu.my/id/eprint/22635/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Health is very essential in everyone's life but to always stay healthy, it becomes a very challenging task especially for the citizens of developing countries. To have a good health, it is important to monitor the physiological parameters such as heart
beat rate/pulse rate, blood pressure, blood oxygen saturation level, respiration rate, temperature and hemoglobin concentration frequently. Nowadays, there are many health care devices that have been developed for measuring physiological parameters but most of them are with limited parameter measurements, a single subject assessment and inconvenient for continuous measurements monitoring due to their contact basis. Furthermore, most of the devices require well-trained health
professionals to operate because the sensors of the devices are to be attached to specific body part for acquiring data. Hence, these drawbacks make the devices suitable to be used at health care centers only. As an alternative approach, this research is focused on extracting physiological parameters through video image processing techniques using ordinary RGB camera. With a recorded video of about 10 seconds, it is possible to analyze multiple physiological parameters simultaneously. The physiological parameters that are extracted in this research
include the vital signs i. e. heart beat rate/pulse rate, blood pressure and blood oxygen saturation level and two other physiological parameters i. e. hemoglobin concentration and skin surface profile. For evaluation of the results, electrocardiogram (ECG), pulse oximeter, oscillometric device and complete blood
count (CBC) test are used to evaluate the results obtained from the developed
video image processing techniques. From the results, it shows that the pulse rate
measurements are quite accurate and within the American National Standard
(ANSI/AAMI EC: 13: 2002) that is ±5bpm or 10% readout error. Besides, the pulse
rate results obtained from the proposed method are able to correlate with ECG,
pulse oximeter and oscillometric device by achieving correlation coefficient of 0.96,
0.97 and 0.95 respectively. In terms of blood pressure measurement, the mean
absolute error and standard deviation for systolic and diastolic pressure from
collected data is 4.45±3.05mmHg and 4.57±3.30mmHg respectively. These values
also fulfill the requirement set by American National Standard (ANSI/AAMI/ISO
81060-2: 2013), which is 5±8mmHg. Furthermore, the correlation coefficient
between the proposed method and oscillometric device is 0.81 and 0.78 for systolic
and diastolic blood pressure respectively. For the blood oxygen saturation level
measurements, the accuracy root mean square error (ARMSi)s 1.26% which is also
able to accomplish the accuracy set in the International Standard ISO 9919: 2005
and ISO 80601-2-61-2011. By comparing the hemoglobin concentration obtained
from the proposed method to the CBC test, the estimated hemoglobin
concentration for the 2 participants are able within the difference of 1 g/dL.
Although there is no standard equipment available for the evaluation of surface
profile in this research, the developed method is evaluated by using the manual
visual inspection approach and the findings of Ondimu and Murase's study. From
the results, it shows that the developed method is feasible to estimate skin surface
profile. In conclusion, the developed video image processing techniques for
extracting multiple physiological parameters simultaneously are very beneficial and
promise high potential due to its non-contact basis, harmless and suitable for
continuous monitoring. Besides, developing the techniques as a smartphone app
would make it more convenient to operate, economical and reduce the white coat
effects, which cause the nervousness when measurements are taken by health
professional. |
|---|