The Development of a Ring-Type Pulse Oximeter Using the Nrf51882 Bluetooth Smart

Table of Contents

Chapter 1 Introduction 2
1.1. Abstract 2
1.2. Background 2
Chapter 2 Literature Review 3

2. 1. Reflected and Transmitted Photoplethysmography 3
   2.2. Factors Affecting Ring-Type PPG Measurements 5
   2.2.1. Light Source and Detector Placement. 5
   2.2.2. Probe Contact Force 8
   2.2.3. Energy Efficiency 9
   2.3. Current Technologies for Ring-Pulse Oximetry 10
   2.3.1. Yu-Chi et al. [10] Ring-Type Pulse Sensor 10
   Block Diagram of the architecture 10
   Critical Analysis of the Design 11
   2.3.2. Park et al. [4] Ring-Type Pulse Sensor 11
   Block Diagram of the architecture 11
   Critical Analysis of the Design: 13
   2.3.3. Huang et al. [6] Ring-Type Pulse Sensor (Multi-Detectors) 13
   Block Diagram of the architecture: 13
   Critical Analysis of the Design: 14
   2.3.4. Kishimoto et al. [11] Ring-Type Pulse Sensor 15
   Block Diagram of the architecture 15
   Critical Analysis of the Design: 15
   2.3.5. Yanchen et al. [12] Ring-Type Pulse Sensor 16
   Block Diagram of the architecture: 16
   Critical Analysis of the Design: 17
   Chapter 3: Conclusion 17
   Chapter 4 Progress report 18
   4.1. Problem Statement 18
   4.2. Methodologies 18
   Chapter 5 References 22

The Development of a Ring-Type Pulse Oximeter Using the Nrf51882 Bluetooth Smart

Beacon Kit and a Pulse Oximeter Circuit
Chapter 1 Introduction

1.1. Abstract
The increasing percentage of the Australian aging population, the prevalence of chronic
diseases, and outbreaks of infectious diseases are some the challenges that the present day
society faces. To address these issues, an active area of health informatics and analytics has
emerged. In particular, acquisition of health informatics through the use of non-invasive
wearable technologies such as, pulse oximeter devices that works together with smartphones.
This paper discusses the current approaches and shortcomings of such wearable technologies
for the monitoring and acquisition of long-term health data, more specifically to the monitor
the heart rate and blood oxygen saturation (������2). Based on my research, I will have to
manufacture a working ring-type pulse oximeter using the nrf51882 Bluetooth smart
beacon kit and a pulse oximeter circuit. The health informatics that I acquire must ultimately
send to a smartphone so the user could read it.
1.2. Background
Pulse oximetry measures continuous arterial blood oxygen saturation, heart rate, and
respiration rate by consecutively turning two light emitting diodes (660nm Red LED, 940nm
Infrared LED) on and off. The light emitted from the diodes is absorbed by tissues, and
the photodetector determines the amount of absorption. Using this information, a
microprocessor can calculate the percentage of oxyhemoglobin by comparing the
concentration of oxyhemoglobin and deoxyhemoglobin in the blood [1]. Blood oxygen
saturation ������2 is calculated as a function of the ratio below:

Figure 1 : Ratio used to calculate ������2 [2]

The general pulse oximeter devices are the finger-type large singular devices with
measurement and display modules integrated into the one unit. However, such finger-type
sensors have their limitations for mobile health monitoring. Below are a few of the reasons
commercial finger-type sensors are ineffective in mobile health monitoring and why there has
been a surge in wearable ring-type sensors [2].
 They are large singular devices that lack effective data management functions
 They have larger power consumption and cost
 They are obtrusive and uncomfortable for long term use
Chapter 2 Literature Review

This literature review will investigate the methods of pulse oximetry applied with different
ring-type pulse sensors. It will explore the current ring-type pulse sensors used in the
biomedical industry and effectively analyze their advantages and disadvantages. Finally, an
assessment will be made on where possible improvements could be done for each ring-type
pulse oximeter device.

2. 1. Reflected and Transmitted Photoplethysmography
   The Photoplethysmography (PPG) pulse wave detection method has two modes of
   operation – transmission and reflectance. In the transmission mode, light transmitted from
   the LED through the tissue. For example, fingers are detected by a photodiode (PD), while,
   in the reflectance mode, the PD detects light that is backscattered from the tissue [1-3].

Figure 2 : Type of PPG Sensor [4]
In 2003, Asada et al. [5] proposed to investigate the effectiveness and limitations of
both the transmittal and reflective illumination methods to monitor heart rate signals and
blood oxygen saturation levels. Their findings were that although both methods are
fundamentally no different from the optic point of view, their functional properties and
performance differs significantly from pressure disturbances, power requirements, and motion
artifacts. They observed that to achieve accurate results, reflective PPG need to be tied down
more securely to the skin surface when compared to transmittal PPG. The findings proved that
once an airgap was created between the reflective PPG and skin surface due to pressure
disturbances such as patient movement, a direct optical path from the LED to PD may be
established consequently leading to saturated measurement. In contrast, transmittal
configurations did not have the same saturation problems, since the LED and PD were placed
on opposite sides of the finger.
To make the ring-type pulse oximeter feasible for long term applications it had to
operate on a coin sized battery pack. Therefore, power requirements had to be taken into
consideration. Despite the superior stability and robustness, transmittal PGG absorbed more
power in comparison to reflectance PPG. Haahr et al. [3] also supported this study, and observed
that the PD in the reflectance PPG mode collects much more of the backscattered light from
the tissue. This feature allowed for a very low forward current to drive the LED, which
consecutively reduced the power consumption on the system. However, the power consumption
problem in the transmittal PPG was solved by turning the LED on for a short time and
sampling the signal within this period. Asada etal. [5] suggest that by attaching an accelerometer
to the ring-type oximeter, we can reduce motion artifacts in both PPG modes of operation.
However, it is further mentioned that given power and size factor requirements, there is no
commercially available product to satisfy such requirements. Figure 3 shows the results obtained
by Asada et al. from an experimental comparison between the two modes of PPG operation.

Figure 3 : Comparison between a continuous transmittal and reflective PPG waveform [5]
Although transmittal illumination methods obtain a relatively better signal than reflective
methods, the measurement sites may be limited. According to Tamura et al. [1] for
transmitted PPG to be effective, “the sensor must be located on the body at a side where
transmitted light can be detected”. Reflectance PGG, however, eliminates the measurement
site restrictions associated with sensor placement, offering a variety of placement sites.
Ultimately the limitation of measurement sites could be taken into consideration while
deciding between the two PPG modes of operation.
2.2. Factors Affecting Ring-Type PPG Measurements
The following are a few key factors affecting PPG measurements that previous
researchers have identified to take into account while designing ring-type oximetry
sensors [1].
2.2.1. Light Source and Detector Placement.
To monitor long-term ������2 health data, the standard fingertip-type sensor will be
inconvenient due to its high sensitivity towards motion. To efficiently overcome this
problem, ring-type sensors are designed. However, due to the complex tissue structure of
the finger-base, shown below in figure 3, the light source and detector placement must be
optimized for accurate ������2 measurements [6].

Figure 4 : Cross-section of the finger base (b) the optical human tissue stimulation when the light source was placed at 0° [6]
To investigate the optimal efficiency of light source and detector placement Huang etal. [6]
conducted an experiment where three types of single detector optical probes were assembled.
The light source was placed at 0° increasing at increments of 15° respectively until it reached
180°. Figure 4 shows the placement of the optical probes conducted in the experiment.

Figure 5 : Different types of optical placements [6]

Huang’s et al. light source and detector simulation results are summarized in the following
table:
R = Ratio obtained from PPG signal used to calibrate and estimate ������2

P-I = Penetrated light to the incident light ratio
Light source & Detector
Placement

R and P-I Ratio Discussions

Light source: 0° – 30°
Detector: 75°- 90°

R ratio is unstable P-I
ratio is very small

Huang et al. reason that obtained
results were the outcome of the
team placing the light source on
the opposite side of the digital
arteries and phalanx that is an
opaque material obstructing the passing light.

Light source: 135°-165°
Detector: 75°- 90°

R ratio stable
P-I ratio is reasonable

Huang et al. observed that by
placing the light source and
detector on the same side of the
digital arteries, a stable R ratio
and reasonable P-I ratio had been
measured. According to their
investigation, the results
mentioned above were obtained
because; the passing light was
able to pass through the digital arteries without being obstructed by the phalanx.

Light source: 135°-165°
Detector: 90°-105°

R ratio is stable
P-I ratio is reasonable

Light
source: 180°
Detector:
105°-120°

R ratio is reasonable
P-I ratio is relatively
small

The explanation for a
reasonable R ratio and relatively
small P-I ratio was established
due to the passing light
travelling a longer distance to
the detector and not passing
sufficiently through the digital
arteries.

Light source: 180°
Detector: 240°- 255°

R ratio is reasonable
P-I ratio is relatively
small

Huang et al. concluded from their simulate on results that the most accurate results
were obtained when they had placed the light source between 135° -165° and the light detector
between 75°- 90° or 90°-105° . To verify these results, ������2 was compared between the
ring-type oximeter and a commercial finger-type pulse oximeter. For the experiment, 10
patients were tested. Each patient was equipped with both types of pulse oximeters and instructed
to hold his/her breath for a period of 70 seconds. The experiment concluded that the correlation
between the ������2 values between both oximeters was 98.26%.

Figure 6 : (a) ������2 measurements from both devices (b) correlation of ������2 between both devices [6]
2.2.2. Probe Contact Force
As mentioned earlier is section2.1, PPG sensors are sensitive to pressure disturbances.
Fundamentally, a valid PPG signal depends on the amount of contact force between the PPG
sensor and the site being measured [1]. Ideally in both transmittal and reflectance modes of
operation the best PPG waveforms are obtained under conditions of transmural pressure. Not
enough pressure and airgaps are created between the reflective PPG and skin surface due to
pressure disturbances such as patient movement, a direct optical path from the LED to PD
may be established consequently leading to a saturated waveform. Furthermore, it is important
to note that too much contact pressure will also result in distorted PPG signals caused by the

occluded artery beyond the PPG probe. Although numerous attempts have been made to
standardize contact pressure in commercial finger-type pulse oximeters no standards have yet
been accepted. In one particular study, Grabovskis etal. [7] tried to determine the optimal
contact pressure. In their study, 5 patients were tested with probe contact pressures ranging
from 0 to 15 kPa. They summarized that the most accurate PPG waveform was acquired when a
contact pressure of 10 kPa (75 mmHg) was applied.
2.2.3. Energy Efficiency
For ring-type pulse oximeters to be practical for long term applications, energy efficiency has
to be taken into consideration in the design process. There are few approaches from
different research teams to optimize energy efficiency. The first consideration taken into
account when controlling power consumption according to Duun et al. [2] is the type of PPG
method used. Asada et al al also concurs with this, both concluding that if power
consumption is the biggest factor to take into consideration, a reflective mode could be the
most energy efficient. The idea being that the PD in the reflectance PPG mode collects much
more of the backscattered light from the tissue. This feature allows for a very low forward
current to drive the LED, which consecutively reduces the power consumption of the
system. As already in section 2.1 turning the LED on fora short time and sampling the signal
within this period, while using energy efficient LEDs would also increase the energy
efficiency. The technique of switching the LED on and off for short time periods was also
used by Park et al. [4] in their approach to tackling the issue of power consumption.
Mendelson etal. [8] mention that by optimizing the amount of light collected by the PD, a very
low power consuming PD could be developed thereby extending overall battery life. Zheng et
al. [9] addresses an innovative technique whereby using inductive powering, the sensors can
be powered partly or completely by RF powering and the battery can be removed. This
method works on the idea that the microcontroller unit is separate to the sensor block. As a
side note I would like to mention that the MIT media lab has built an unobtrusive device
which retrieves energy from the heel-strike of a walking user [9]. This device could

potentially power wearable PPG devices. Below is the timing diagram used by Park et al. [4]
to control power consumption.

Figure 7 : Timing diagram to control the ring sensor [4]

2.3. Current Technologies for Ring-Pulse Oximetry
Below are the hardware and architecture of 5 different types of ring-type pulse oximeters
along with their advantages, disadvantages, and key attributes.
2.3.1. Yu-Chi et al. [10] Ring-Type Pulse Sensor
Hardware Components:

1. Physiological Sensor: Transmittal ring-type sensor designed by Sinopulsar
   Technology Inc. The sensor has an inbuilt RFID.
2. Wireless Communication method: Bluetooth v2.1 module
3. Monitoring Interface: Smart phone with Windows operating system

Block Diagram of the architecture

Figure 8 : Block diagram of architectureFigure 9 : RFID ring-type sensor

Critical Analysis of the Design
PPG
Mode

Advantages Disadvantages Key Attributes

Transmittal

The system uses
minimum energy for
data transfer between
the RFID reader and
smartphone.
Applying the required
amount contact
pressure between the
detector and
measurement site is
not an issue.

RFID reader is too
obtrusive for long
term management of
health data.
The Accuracy of the
ring-type sensor has
not been validated
against commercial
finger-type oximeters.

Data acquired by the
smartphone is
transferred to a remote
location through
mobile
communication to
alert health officials.

2.3.2. Park et al. [4] Ring-Type Pulse Sensor
Hardware Components:

1. Physiological Sensor: Reflectance ring-type sensor
2. Wireless Communication method: ZigBee protocol
3. Monitoring interface: PC
   Block Diagram of the architecture

Figure 10 : Block diagram of architecture

Figure 11 : RFID Ring-type sensor

Critical Analysis of the Design:
PPG
Mode

Advantages Disadvantages Key Attributes

Reflectance

RFID reader is
wearable for
long term
management of
health data.

Power saving
mode which
switches the
LED on and
off for short
time periods.

The ring sensor must send 512
bytes to calculate the heart rate.
Because the Zigbee standard
limits the payload the ring sensor
must transmit the data 8 times.
The Accuracy of the ring-type
sensor has not been validated
against commercial finger-type
oximeters.
Applying the required amount
contact pressure between the
detector and measurement site is
an issue.
PPG signal was significantly
affected from motion artifacts.

Ring sensor, Power
block and RFID
microcontroller
block incorporated
into an unobtrusive
package.

2.3.3. Huang et al. [6] Ring-Type Pulse Sensor (Multi-Detectors)
Hardware Components:

1. Physiological Sensor: Reflectance ring-type sensor
2. Wireless Communication method: Bluetooth v2.0 module
3. Monitoring interface: PC
   Block Diagram of the architecture:

Figure 12 : (a) Block diagram of architecture (b) RFID Ring-type sensor

Critical Analysis of the Design:
PPG
Mode

Advantages Disadvantages Key Attributes

Reflectance

Ring sensor, Power block and
RFID microcontroller block
incorporated into a semi-
unobtrusive package.
Optimized detector and light
placement for accurate PPG
waveforms and efficient power
consumption.
The Accuracy of the ring-type
sensor has been validated
against commercial finger- type
oximeters.
Correlation between the ������2
values between both oximeters
was 98.3%.

Applying the
required amount
contact pressure
between the
detector and
measurement site is
an issue.

PPG signal was
significantly
affected from
motion artifacts.

Used multiple
detectors which
when compared to
single detectors
produced more
stable ������2 results.
However, by adding
multiple detectors
battery life was
decreased.

2.3.4. Kishimoto et al. [11] Ring-Type Pulse Sensor
Hardware Components:

1. Physiological Sensor: Reflectance ring-type sensor
2. Wireless Communication method: Telephone line or internet
3. Monitoring interface: PC
   Block Diagram of the architecture

Figure 13 : Diagram of architecture

Figure 14 : RFID Ring-type sensor
Critical Analysis of the Design:
PPG Mode Advantages Disadvantages Key Attributes

The Accuracy of the
ring-type sensor has
been validated against
commercial finger-
type oximeters.

Applying the required amount
contact pressure between the
detector and measurement site
is an issue resulting in
distorted measurements.

Not only does the
ring-type sensor
relay
data it also has

Reflectance

Correlation between
the ������2 values
between both
oximeters was 82%.
The ring-type sensor
has an inbuilt
display.

A lot measurement lost due to
Multiple battery rundowns
during testing. Needs a more
stable power supply.
Acquired data is sent
monitoring interface through a
telephone line or internet.

inbuilt display for
heart rate and ������2.

2.3.5. Yanchen et al. [12] Ring-Type Pulse Sensor
Hardware Components:

1. Physiological Sensor: Transmittal ring-type sensor
2. Wireless Communication method: Bluetooth v4.0 module
3. Monitoring interface: Mobile phone running iOS
   Block Diagram of the architecture:

Figure 15 : Block diagram of architecture

Figure 16 : Smartphone APP display interface

Critical Analysis of the Design:
PPG
Mode

Advantages Disadvantages Key Attributes

Transmittal

The Accuracy of the ring-
type sensor has been
validated against
commercial finger- type
oximeters.
Correlation between the
������2 values between both
oximeters was 98.5%.
Saves power and reduces
circuit complexity by
implementing the data
processing into the
smartphone.

Signal disposal module
is too obtrusive for long
term health
management.
PPG signal was
significantly affected
from motion artifacts.

Compared to the
other ring-type
sensors mentioned
this is very simple yet
reliable.

Chapter 3: Conclusion
As noted by Asada et al. [5], the transmittal and reflective illumination methods can be
used for monitoring heart rate signals albeit with a few difficulties like motion artifacts, high
power requirements, and pressure disturbances. Some of the problems that are expected in the
course of designing the device are probe contact force and energy consumption. Grabovskis etal.
[7] resolve the problem of contact force by suggesting a contact pressure of between 0 and 15 kPa.
The wearable ring-type pulse oximeters must have a small battery size battery to reduce their size
and weight. The type of PPG used is the chief measure of increasing the energy efficiency of the
device [2]. Park et al. [4] suggest the technique of switching the LED on and off in short periods
to save power. On the other hand, Mendelson et al. [8] suggest the option of optimizing the
amount light collected by the PD by developing a very low power consuming PD.

Chapter 4 Progress report
4.1. Problem Statement
Using the nrf51882 Bluetooth smart beacon kit as a microprocessor and signal disposing
module, I will manufacture a working ring pulse oximeter that sends processed oximetry
data to a smartphone, which is in turn interpreted by the user. The challenges I am faced
with are as follows:
 Power consumption and Component selection: The entire system has to operate
on a coin sized battery pack of 3.3V if it’s to be practical for the long term
management of health-data. The photodiode must have an in inbuilt amplifier and
filter (1-3 Hz), operate within a minimum operating voltage of 3.3V, and a
viewing angle of (18°-35°). The light emitting diodes must emit light at a
wavelength of 940nm and 660nm respectively, operate with a forward voltage less
than 3.3V and a maximum forward current of 150mV. The viewing angle is to be
determined by testing various types of LEDs.
 Circuitry: Biasing design of the LED’s, small and compact circuit design
 Data collection and processing: Testing the collection and processing of data on
the smartphone or Bluetooth beacon depending on the methodology that yields the
most accurate results and least power consumption.
 Building: Combining all the components into a small unobtrusive and wearable
technology.
4.2. Methodologies
The solution to choosing the correct components and circuitry is through an iterative
build-measure-learn feedback loop. By comparing the waveform obtained from the output of
the ring-type pulse oximeter with the expected waveform from a commercial finger-type
pulse oximeter, I can decide which setup and component configuration yield the most
accurate results. The Data collection and processing procedure will be broken down into two
stages, the first phase being the acquisition of data using the Arduino board and processing

the obtained data using Matlab. The second phase will be using the same code from stage one
on the nrf51882 Bluetooth smart beacon and transmitting the data to the smartphone through
the beacons inbuilt RFID technology.
My approach this semester was to implement a working pulse oximeter prototype. The
idea is to minimize the overall design size of the pulse oximeter at every iteration. I based
my pulse oximeter off the following circuit design and architecture below.

Figure 17 : Main block diagram of the pulse oximeter system

Figure 18 : Circuit with an optical sensor with red and IR LED and phototransistor

Figure 19 : Filtering and amplifying circuitry

Figure 20 : First pulse oximeter prototype

Components:
 Infrared LED SFH4110, red LED 660nm, SDP8406-003 silicon photo-transistor
 Dual op-amp LM358
 Capacitors 10µF X 2, 470nF X 2
 Resistors 4.7KΩ, 1 KΩ, 100 KΩ, 38 KΩ, 2 X 100 Ω
 Diodes IN914 X 2
 Breadboard
Figure 6 below shows the initial waveform I acquired from testing. Figure 7 shows the
waveform after I optimized the housing of the sensor head. Just by blocking out background
light I was able to obtain a clear heart signal.

Figure 21 : Detected PPG waveform

Figure 22 : Detected PPG waveform after optimizing the sensor head
By measuring the peak-peak distance measurements I was able to calculate my heart
rate to be 83 bpm. I compare this to the inbuilt pulse oximeter app on my smartphone that
measured my heart beat to be 85bpm. My first goal for next semester is to incorporate my
circuit design into a small breakout board with a surface mounted op-amp for the filtering and
amplification phase. From there every prototype will become smaller in size.
Honours/ Thesis Part B
NO Tasks Sesmester 1 2016 (Weeks)
1 2 3 4 5 MIDSE

MBREAK 6 7 8 9 10 11 12 13

1 Ring Pulse
Oximter Research 2 nrf51822
bluetooth beacon 3 -Optical
Sensor Head 4 -Electronic
Signal Process 5 -Building
and Testing 6 -Data
Collection/Processing 7 Ring
8 -Design
9 -Manufacture
10 App Desin and
Build 11 Draft thesis
12 Presentation
13 Final Thesis

Chapter 5 References

[1] T. Tamura, Y. Maeda, M. Sekine, and M. Yoshida, “Wearable photoplethysmographic
sensors—past and present,” Electronics, vol. 3, pp. 282-302, 2014.

[2] S. B. Duun, R. G. Haahr, K. Birkelund, and E. V. Thomsen, “A Ring-Shaped Photodiode
Designed for Use in a Reflectance Pulse Oximetry Sensor in Wireless Health
Monitoring Applications,” Sensors Journal, IEEE, vol. 10, pp. 261-268, 2010.

[3] R. G. Haahr, S. Duun, K. Birkelund, P. Raahauge, P. Petersen, H. Dam , et al., “A
Novel Photodiode for Reflectance Pulse Oximetry in low-power applications,” in
Engineering in Medicine and Biology Society, 2007. EMBS 2007. 29th Annual
International Conference of the IEEE, 2007, pp. 2350-2353.

[4] S. M. Park, J. Y. Kim, K. E. Ko, I. H. Jang, and K. B. Sim, “Real-time heart rate
monitoring system based on ring-type pulse oximeter sensor,” Journal of Electrical
Engineering & Technology, vol. 8, pp. 376-384, 2013.

[5] H. H. Asada, P. Shaltis, A. Reisner, R. Sokwoo, and R. C. Hutchinson, “Mobile
monitoring with wearable photoplethysmographic biosensors,” Engineering in
Medicine and Biology Magazine, IEEE, vol. 22, pp. 28- 40, 2003.

[6] C.-Y. Huang, M.-C. Chan, C.-Y. Chen, and B.-S. Lin, “Novel Wearable and Wireless
Ring-Type Pulse Oximeter with Multi-Detectors,” Sensors (Basel, Switzerland), vol.
14, pp. 17586-17599, 09/19

[7] A. Grabovskis, Z. Marcinkevics, U. Rubins, and E. Kviesis-Kipge, “Effect of probe
contact pressure on the photoplethysmographic assessment of conduit artery

stiffness,” Journal of Biomedical Optics, vol. 18, pp. 027004-027004, 2013.

[8] Y. Mendelson and C. Pujary, “Measurement site and photodetector size considerations
in optimizing power consumption of a wearable reflectance pulse oximeter,” in
Engineering in Medicine and Biology Society, 2003. Proceedings of the 25th Annual
International Conference of the IEEE, 2003, pp. 3016-3019 Vol.4.

[9] Z. Ya-Li, D. Xiao-Rong, C. C. Y. Poon, B. P. L. Lo, Z. Heye, Z. Xiao-Lin, et al.,
“Unobtrusive Sensing and Wearable Devices for Health Informatics,” Biomedical
Engineering, IEEE Transactions on, vol. 61, pp. 1538- 1554, 2014.

[10] W. Yu-Chi, C. Pei-Fan, H. Zhi-Huang, C. Chao-Hsu, L. Gwo-Chuan, and Y. Wen-
Ching, “A Mobile Health Monitoring System Using RFID Ring-Type Pulse Sensor,”
in Dependable, Autonomic and Secure Computing, 2009. DASC ’09. Eighth IEEE
International Conference on, 2009, pp. 317-322.

[11] A. Y. A. Kishimoto, O. Tochikubo, K. Ohshige, and A. Yanaga, “Ring-Shaped Pulse
Oximeter and Its Application: Measurement of SpO2 and Blood Pressure during
Sleep and During Flight,” Clinical and Experimental Hypertension, vol. 27, pp. 279-
288, 2005/01/01 2005.

[12] D. Yanchen and L. Jian, “Design of Noninvasive Pulse Oximeter Based on Bluetooth
4.0 BLE,” in Computational Intelligence and Design (ISCID), 2014 Seventh
International Symposium on, 2014, pp. 100- 103.