Advanced Wearable Technology for Cardiovascular Health

DESIGN, DEVELOPMENT, AND FABRICATION OF WEARABLE PHOTOPLETHYSMOGRAPHY FOR CARDIOVASCULAR MONITORING

Abstract

Improvement in the area of the technology of sensors, ubiquitous connections and computing solutions has covered the path for the wearable’s development for random monitoring of physiological parameters. The rate of heart monitoring by the accuracy of clinical reflectance signals of “photoplethysmography” from the wrist is the most challenging. The device executes upon a low-energy M4 microcontroller of ARM cortex and BLE. rate of heart calculated through the worn device of the wrist has existed against the rate of heart calculated by “Masimo Radical”. The study’s validation was hosted in a setting of hospital for a 1-month duration.

Keywords: “photoplethysmography”, “rate of heart”, and “accelerometer”, “wireless”.

Table of Contents

Introduction 4

Methods and materials 5

Results 10

Carrier design 12

Conclusion 12

References: 13

Introduction

Nowadays, the population is increasing, and the demand for electronic wearable devices has also increased simultaneously to measure physiological conditions of heart-related issues. Different techniques that are used for sensing, and continuous and instant measurement of physiological parameters hold an important role in identifying certain health conditions. Recently, for people who have a high chance of encountering heart attacks and strokes, small sensor systems that are wearable have been designed. Wearable systems and devices are used on a large scale to collect various information regarding muscle activity and the sleep cycle of an individual in their daily life. There are different forms of wearable technology that have been discovered in recent years such as integrated clothing, body attachments, and accessories to name a few. Photoplethysmography sensors are the best examples that use infrared light in order to measure changes in the flow of blood at the skin surface level [1]. Arterial Oxygen Saturation detection, blood pressure monitoring, arterial aging, and microvascular flow of blood are observed with the help of PPG or Photoplethysmography Sensors. These sensors can also be developed as skin-on interfaces by implementing computing devices and other sensors directly on the surface of the body. Wearable devices have the ability to allow continuous measurement of different physiological data. Electronic devices that are flexible can be placed on the skin to detect different signals using high sensitivity. This paper provides a structured review of PPG technology’s latest development that is used for clinical applications [2]. This sensor’s applications and operational mechanisms are first presented, and then different developments and progress in materials to get the stable performance of light sensing are put forward. Finally, the current limitations and challenges that PPG sensors are encountering are highlighted.

Problem Statement

This study is performed to highlight different limitations electronic wearable devices are facing during their operation in everyday life so that this can be rectified in future studies. Wearable devices often cause skin irritation due to friction between the skin and the back of the device. This happens due to the system’s tracker that does not allow sweat to evaporate from the skin thereby increasing rashes, red spots, and inflammation.

These devices also have low durability during operation due to repeated deformation. Apart from that, these sensors consume high power when used for a long period of time. Photoplethysmography Sensors are basically used for measuring blood flow change and it uses infrared light that consumes high power. This happens due to heavy and hard device architecture that is commonly used in semiconducting materials that are photoresponsive. This makes it very important to make the selection of materials that are used for sensing and skin integration of Photoplethysmography Sensors to obtain various functionalities and mechanical properties. Furthermore, this paper provides a systematic review of Photoplethysmography Sensor’s latest technological development and their usage in the clinical industry. Finally, certain limitations and challenges that PPG or Photoplethysmography Sensors are facing during their operation in the healthcare system are being put forward.

Methods and materials

Reflectance PPG

“Reflectance photoplethysmography”, the tissue of the skin is lightened through a “light emitting diode”. In addition, the “light intensity reflected” is calculated by the support of a photo detector. Moreover, this structure is presented in figure 1 and this is the common modality calculation adopted in the device. The rate of the heart can be blend through analysis of the contrast in the reflected intensity of light.

Figure 1: light reflectance from skin and vessels

(Source : )

According to figure 1, it has been seen that five major elements are involved here and these are “the epidermis”, “papillary dermis” and “blood vessels”. In addition, “hypodermis” and “reticular dermis” are also enlisted in this figure 1. Apart from that, in this figure, LED and PD have a major role that has been displayed here.

Accurate rate of heart and reliable computation through reflectance photoplethysmography situated at the wrist of the body. This acted as the most challenging for the wrist. On the other hand, the content of Melanin in the skin and the structure of tissue affect the perfusion.

Design

The protocol of the stage block diagram about the heart rate of the optical monitor is presented in figure 2. The “wrist-worn optical HRM” has two “PCB stacked architectures”. In addition, the PCB of the sensor board houses the PD and LED required to monitor the heart rate.

Figure 2: Block diagram of the system level

 

Abovementioned figure 2 has been presented with the block diagram related to the “microcontroller context M4”’s branches and working procedures. On the other hand, “power supply” has three parts and these are “sensor PCB”, “ AFE” and “Microcontroller context M4”.

On the other hand, the twin architecture of PCB implemented in the HRM of optical has been presented in figure 3.

Figure 3: the sensor board and stacked motherboard by the battery in among sensor detecting green PD, Green LED along with capacitive sensor of touch.

The worn of wrist device is made to use conjunction through a tablet. In addition, this device communicates wirelessly by low-energy of Bluetooth protocol.

1. Motherboard PCB

The control and processing operations about the worn of wrist devices related to wearable’s are managed by a low-energy “ARM Cortex M4 microcontroller”. The HRM of wearable has 3 axes of an ultra-low energy MEMS accelerator of digital by a common assumption of current about 2μA and it is used in removing motion artifacts. The outcomes of measurement are gathered in a house of flash memory in the PCB section of the motherboard. The device meets the outcome contained in the memory portion by the device of the gateway [4]. In addition, this entire process happened through the using protocol of BLE. Furthermore, the move of information to the device of the gateway is later pressed onto the cloud due to analysis of the data. Buzzer and e-link display and two press buttons are served in the optical of HRM to activate the “Human Machine Interface” by the user and device. The device is provided through a 200mAH battery of Li-Po. Moreover, this battery can execute for 48 hours on a charge by measurements of periodic.

1. Analog front end

The front of the analog end is required for the rate of heart computation to be accommodated on the top of the PCB of the motherboard. In addition, some items are included that are presented in figure 4. Some elements are related to “trains impedance Amp”, “DC removal” and “programmable gain Amp”. The LEDs are moved through MOSFET. On the other hand, control intensity is gained through enlisting an input of PWM at the gate [2]. The PD is associated with an amplifier of trans-impedance that transforms the Ampere of sub-micro current. In addition, the result of Triamp is taken to a subtraction stage of DC that conducts down the offset of DC in the sign.

Figure 4: Front end of analog

The removal of the DC level is maintained through a “programmable Gain Amplifier”. In addition, signal acquisition is covered by a 16-bit successive at the range between analog and digital converters.

Algorithm

A representation of a flow chart about the implementation algorithm is attached in figure 5. In this case, the algorithm utilized in the tools for computing the heart rate. Apart from that, this algorithm has three several parts such as “sensor placement check”, “ PGA gain set” and “signal acquisition”.

Check of sensor placement

Photo detector’s proper placement and assembly of LED on the surface of the skin by the tight air gap. In addition, this is generated with touch sensors of capacitive. This capacitive touch sensor is situated under the clock due to monitor continuously skin.

PGA gain test and intensity

In this case of operation, a PID controller has utilized in sets of the firmware the green LED’s intensity. This is completed to confirm that the front of the analog end is tuned to obtain exact signs of PPG for aspects by skin tone [3]. The intimates of device the public to dress in the gadget accurately in case intensity cannot be fixed even 3 iterations.

Signal acquisition

Exact contact is confirmed acceleration and PPG sing acquisition occurs randomly in the background utilizing DMA. present implementation implies 60 rates of sampling per second due to PPG. the common rate of heart over a time of about 10 seconds is calculated through acquiring PPG’s 600 samples.

Rate of heart computation

A plot of the waveform due to signals of raw PPG covered through the help of HRM and accelerometer data filtering is presented in figure 5.

Figure 5: normalized section’s raw PPG waveform duplicated by effects of motions

Artifacts of motion is reduced by utilizing the sum vector which is filtered in the previous about 3-axis data of acceleration. In addition, in this study one equation is enlisted that is mentioned below.

Figure 6: equation of 3-axis data of acceleration

Figure 6: implementation of the algorithm

According to figure 6, it has been seen that some steps such as “sensor placement check” and “signal quality calculation” are included in the entire algorithm. In addition, the items are “device as well as peripheral initialization”, “sensor check of placement”, and “600 samples of signal Acquisition”. The total waveform of the vector is lower pass purified by a frequency related to cutoff at 0.4Hz for getting waveform of motion.

The rate of the heart is calculated from those portions related to the PPG waveform in the place of data about the accelerometer. In addition, it does not cross the threshold of motion. The threshold of motion was measured after completing the process of motion analyzing [4]. The rate of heart is calculated with the help of time domain and frequency analysis to meet the average values.

Clinical validation

In order to check the reliability and accuracy about the HRM of optical within a setting regarding clinical aspects. The device pieces of information and clinical sample system were displayed in the review. Moreover, the committee of ethics approval was gathered prior to the trials of clinical. The clinical paper endpoints are enlisted below. The first one is the “average rate of heart’s accuracy calculated over a 10 seconds duration by artifacts of motion.”. The second one is to “analyze and understand the skin tone effects in the calculated value of heart rate”.

Results

The won of wrist optical HRM was measured in the setting of clinical on 256 subjects. In addition, this operation is executed by giving 256 subjects. In 256 subjects, 116 are enlisted under the female category and 140 are included under the list of males. Sometimes, the worm of the wrist optical rate of the heart monitor detected tablet by a message detecting that the rate of the heart monitor was not tattered exactly [6]. The mean value of error in the rate of the heart is calculated through the device during constraining by measuring the rate of heart. In order to examine the agreement’s degree among the readings gathered from the rate of the heart monitor calculation. In addition, this scenario has been presented in figure 7.

Figure 7: heart rate’s Bland Altman calculated on the wrist by the rate of optical heart Massimo radical-7 and monitor

Figure 8: clinical paper pieces of information analyzed based on age groups

According to figure 8, columns are involved and these are “age group”, “subject count” and “standard definition”, “mean value”. Moreover, the subject count is situated at 256 and the standard value exists at 5.29. In addition, the mean value stands at -0.39. The remarks detected the skin tones of about 256 subjects, and 54 are selected as fair. Furthermore, 181 were counted as moderate level and the rest 21 were counted as dark. In form of the figure 10, it has been seen that the exact error within the rate of heart calculated for single by single in categories in a whisker plot of comparison and a box.

Figure 9: constraint about the exact error in the rate of heart calculation for different skin tones.

According to figure 9, it has been observed that this operation defined the different kinds of skin tones such as dark, and fair. The top and button of the box detect 25th along with the 75th percentile [7]. The whisker ends present the lower level datum in 1.5 IQR about the low-level quartile. The error of the mean from the plot due to skin which belonged to the skin is 1.04 bpm. Therefore, this study has forecasted the results section by inputting the accurate value related to the operation.

Carrier design

I have faced lots of barriers such as time constraints, and lack of data during the time of collecting data. On the other hand, insufficient data played a major role as an issue about collecting information related to this topic. In addition, I have received error data about this clinical technology. Apart from that, I have received a huge amount of support from my supervisors and my friends during information collection. Therefore, I want to provide exact acknowledgement to all my supporters. In the time sphere of grasping the data, I have not been able to collect accurate information regarding this project. At that time, my supervisors provided me with the required materials for demonstrating this project with full of information. Thus, I have made this project by grabbing a bulk amount support besides of facing constraints.

Conclusion

According to the study focused on the entire paper presented an error in the calculation related to heart rate. As per the operation, it has been founded that the dark skin heart rate value is lower than the exact heart rate. Artifact of motion removal through eliminating the portions of the signal that are corrupted. The tool could be executed randomly by recent implementation along with optimization of power during 40 hours. Multiple wavelengths used for the rate of heart calculation might develop a quality due to dark skin tone subjects. Thus, this research has tried to demonstrate deeply about the selected context. Apart from that, this entire process has presented the accurate values of all relevant data related to enlisting all tables.

References:

1. Tamura, T., Maeda, Y., Sekine, M. and Yoshida, M., 2014. Wearable photoplethysmographic sensors—past and present. Electronics, 3(2), pp.282-302.
2. Sun, Y. and Thakor, N., 2015. Photoplethysmography revisited: from contact to noncontact, from point to imaging. IEEE transactions on biomedical engineering, 63(3), pp.463-477.
3. Preejith, S.P., Alex, A., Joseph, J. and Sivaprakasam, M., 2016, May. Design, development and clinical validation of a wrist-based optical heart rate monitor. In 2016 IEEE International Symposium on Medical Measurements and Applications (MeMeA) (pp. 1-6). IEEE.
4. Dubey, H., Constant, N. and Mankodiya, K., 2017, July. RESPIRE: a spectral kurtosis-based method to extract respiration rate from wearable PPG signals. In 2017 IEEE/ACM International Conference on Connected Health: Applications, Systems and Engineering Technologies (CHASE) (pp. 84-89). IEEE.
5. Constant, N., Wang, T. and Mankodiya, K., 2015, March. Pulseband: A hands-on tutorial on how to design a smart wristband to monitor heart-rate. In 2015 IEEE Virtual Conference on Applications of Commercial Sensors (VCACS) (pp. 1-3). IEEE.
6. Hajir, M. and Khan, T.S., 2014. Development of wearable pulse oximetry for telehealth monitoring system. Journal of Electrical Engineering, 14(4), pp.6-6.
7. Lee, C.S., Wu, C.Y. and Kuo, Y.L., 2017. Wearable bracelet belt resonators for noncontact wrist location and pulse detection. IEEE Transactions on Microwave Theory and Techniques, 65(11), pp.4475-4482.