View of Development of MEMS accelerometer based real time data acquisition prototype for COVID-19 patient breath monitoring and assessment

Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 5190-5197

Research Article

5190

Development of MEMS accelerometer based real time data acquisition prototype for

COVID-19 patient breath monitoring and assessment

Kalyan Dusarlapudi

1

_{, Narasimha Raju K}

2

_{, Preeti M}

3

_{, Koushik Guha}

4

_{, Venkata Siva}

Kumari Narayanam

5

1_{Assistant Professor, Department of Electrical and Electronics Engineering, Koneru Lakshmaiah Education} Foundation, Vaddeswaram, A.P, India.

2_{Professor, Department of Electrical and Electronics Engineering, Koneru Lakshmaiah Education Foundation,} Vaddeswaram, A.P, India.

3_{Department of Electronics and Communication Engineering, Koneru Lakshmaiah Education Foundation,} Vaddeswaram, AP, India,Electronics and communication engineering department, National Institute of technology, Silchar, Assam, 788010, India

4_{Electronics and communication engineering department, National Institute of technology, Silchar, Assam,} 788010, India.

5_{Assistant Professor, Communication and soft skills department, Koneru Lakshmaiah Education Foundation,} Vaddeswaram, A.P, India.

Article History: Received: 10 January 2021; Revised: 12 February 2021; Accepted: 27 March 2021; Published online: 28 April 2021

Abstract:In late 2019, coronavirus-2, a dangerous human pathogen that causes the disease coronavirus disease 2019, arose as

a severe human pathogen (COVID-19). Acute respiratory failure, which is similar to acute respiratory distress syndrome, is the most frequent clinical presentation of extreme COVID-19. COVID-19 includes diseases of the airway, lung parenchymal, pulmonary vascular, and respiratory neuromuscular systems. This essay summarises what is learned about the impact of coronavirus-2 infection on various areas of the respiratory system, provides insights into the basic pathology of respiratory disease, and highlights current and prospective translational and clinical research. This paper discusses economic and reliable respiratory monitoring systems using prototype popular embedded controller and built-in smartphone MEMS accelerometer. The hardware is tested and implemented on healthy, and Covid-19 affected persons for respiratory data, and Correlation has been made among them. One class SVM(Support Vector Machine) learning algorithm further processes the data collected from the sensor. The project primarily aims to detect anomalies in the respiratory data set to assess the rate of respiration by matching the data sets of a healthy individual and a COVID infected person.

Keywords: COVID-19, MEMS Accelerometer, SVM(Support Vector Machine), Coronavirus

1. Introduction

The COVID-19 pandemic has prompted a surge in remote patient control. Much of the patients who tested (or are accused of testing) positive for COVID-19 are in self-isolation at home as of the time of writing this report. Healthcare staff, personal protection devices, hospitals, and automatic ventilators in intensive-care units are all in limited supply, even in the most robust healthcare facilities (ICU). As a result, new medical options, such as remote patient care, are required. As a result, new regulations have been implemented to encourage the advancement of tracking systems, thus providing favourable conditions for improving remote monitoring of certain vital signs that have previously been ignored. This is particularly true for respiratory rate (RR), which is currently under-recorded considering its importance in COVID-19.

According to current COVID-19 recommendations, resting RR can be measured to help with triage, diagnosis, and prognosis, as well as as a criteria for ICU entry and early detection of COVID-19 patient decline. A resting value of RR > 30 breaths/min, according to the World Health Organization, is a vital indication for the diagnosis of serious pneumonia in adults, although the cut-off value for children varies depending on age (World Health Organization, 2020). RR values are used at triage to help refer patients to various groups and make recommendations on whether or not to use supplemental oxygen.

Invasive mechanical ventilation is needed for many COVID-19 patients with acute respiratory failure to prevent further impairment in gas exchange, respiratory muscle weakness, organ failure, and death. Invasive mechanical ventilation for long periods of time (up to several weeks) has been recorded. 90 Although the effects of SARS-CoV-2 and sustained mechanical ventilation on respiratory muscle development and operation are uncertain, they may be harmful and clinically significant. The diaphragm is the primary inspiratory muscle, and it tends to be more affected by critical illness and mechanical breathing than peripheral muscles. Diaphragm feature should be improved[7].

Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 5190-5197

Research Article

5191

Fig.1. Remote Recording of Respiratory data with variety of monitoring technologies

COVID-19 patients will benefit from technological tools for remote RR control. Figure 1(A) A smart device's built-in camera may be used to monitor RR from respiratory-induced chest wall vibrations or superficial changes in a seated patient's face perfusion. Figure 1 (B) A smartphone's built-in microphone can be used to capture RR from the patient's breathing sounds. Figure 1 (C) A patient's breathing-related chest wall motions may be recorded using an instrumented mattress for continuous RR tracking. Figure 1 (D) The modulation of transmitting signals by respiratory-related chest motions may be used to record RR values using radio wave or Wi-Fi signal sources and receivers (no body sensors are needed). Figure 1 (E) Even during everyday operations, a smart fabric (e.g., a strap with conductive strain sensors) may be used to constantly record RR from respiratory-related periodic changes in chest wall circumference (e.g., walking).[7]

2. Literature review

The literature papers[1] present a device architecture for continuous and long-term respiratory rate monitoring in a wearable form factor with remote monitoring capabilities. The computer is powered by a low-power ARM Cortex MO microcontroller and has a built-in Bluetooth Low Energy module. For dependable efficiency, the computer uses a respiratory rate computation algorithm with motion artefact rejection. The proposed modality's machine architecture, application configuration, algorithm implementation, and experimental validation are all presented.

In [2] it was suggested that researchers check Medline, Embase, CINAHL, and reference lists for studies that registered the heart rate or respiratory rate of healthy children from birth to 18 years. They created centile charts for heart rate and respiratory rate in relation to age using non-parametric kernel regression. Current reference ranges were compared to those obtained from our centile maps.

The new method was suggested[3] by the writers M. R. Ambedkar and S. Prabhu. An EEMD approach is proposed to extract the necessary respiratory information from the PPG signal. This algorithm works well for signals that are not stationary. According to this report, the results obtained by EEMD are significantly better than those obtained by EMD. The EEMD device efficiently collects respiratory image. The system achieves an overall precision of 97 percent. As a result, the obtained results show that EEMD is efficient. The algorithm was written in MATLAB R2013 [5].

3. Methodology

Several techniques exist for determining the rate of respiration. A common technique is to use a chest strap or a pneumographic device dependent on impedance. The capnograph is a device for calculating respiratory rate that is often used in hospitals. Plethysmography is another tool that has been proposed for determining respiratory rate. Mechanical instruments including a magnetometer, strain gauge, or gas pressure sensor are commonly used to calculate the diaphragm's displacement. The rate of respiration can also be calculated using a photoplethysmogram or electrocardiogram waveform analysis. Such methods are less precise, and others are vulnerable to motion corruption. Telemedicine is a branch of medicine that employs technologies to treat patients who are spread out over a large geographic area. A telemedicine device allows a specialist in one area to treat a patient in another.

Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 5190-5197

Research Article

5192

In this paper, M E M S accelerometer is used to determine acceleration (using a non-contact device) in three mutually perpendicular directions (X, Y and Z). A Bluetooth module and a Triaxial Accelerometer ADXL335 sensor are used to monitor accelerations at a location where measuring accelerations is usually impossible due to the inaccessibility of a narrow region where three accelerometers cannot be reached.

4. Design of DAQ system 4.1 Block diagram

The proposed hardware to extract the respiration data of healthy person and COVID19 effected is based on various embedded processors. For compactness Arduino Nano module is selected and integrated with Accelerometer and 433MHz wireless transmitter and receiver. Fig2 shows the components needed to form data acquisition system.

Fig.2. Block diagram of Wireless DAQ system

Fig.3. Components of Wireless DAQ system

The ADXL335 is a monolithic three-axis acceleration measuring unit. The minimum measurement range of the ADXL335 is 3 g. It requires a polysilicon surface-micromachined sensor and signal conditioning circuitry to incorporate an open-loop acceleration measuring architecture. As output signals, analogue voltages equal to acceleration are used. The accelerometer can measure the static acceleration of gravity as well as dynamic acceleration caused by rotation, shock, or vibration in tilt-sensing applications. To quantify the structure's deflection, a differential capacitor with separate fixed and rotating mass plates is used. The plates are moved by 180° out-of-phase square waves. Acceleration allows the differential capacitor to unbalance, resulting in a sensor signal with a relative amplitude to acceleration. Using phase-sensitive demodulation methods, the amplitude and trajectory of the acceleration is then determined.

Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 5190-5197

Research Article

5193

Fig.4. MEMS Accelerometer module

The sensor data is analysed using the Arduino Microcontroller after the ADXL335 Accelerometer is attached to the normal and diseased people's abdomens. PLXDAQ programme reads data from the serial monitor of the Arduino IDE and transfers it to the local computer.The csv files contain information from the PLXDAQ software. The aim of the project is to combine both data sets. The first set of data comes from a good person, and the second set comes from someone who is ill. This data sets are then used to compare the regular and diseased data sets using various algorithms.[6]

5. Experimental Setup:

The ADXL335 accelerometer circuit is simple and does not need any complicated parts or wiring. It can be directly attached to an Arduino. The measures for connecting the ADXL335 accelerometer to an Arduino Uno are as follows: Attach the VCC pin to the Arduino's 5V pin, the GND pin to the Arduino's Ground pin, and the X,Y, and Z data outputs to the analogue pins A0, A1, and A2 on the Arduino. The ADXL335 calculates acceleration along the X, Y, and Z axes and outputs analogue voltage equal to the acceleration. This voltages can be processed by microcontrollers by translating them to digital signals using ADC. The Wireless modues 433MHz RF transmitter and received are integrated to the Arduino for wireless data transfer of respiration data[10].

The Parallax Data Acquisition Tool (PLX-DAQ) programme free add-in for Microsoft Excel can be used to conveniently link the Arduino to Excel. ages eight and eleven The Arduino would need to be installed with the Arduino software before it could be connected to Excel (IDE). Click the PLX-DAQ spreadsheet icon after uploaded the Arduino code. Select the Arduino port, search the Download Data box, and press Link. To see real-time plotting, one should add a line graph.

Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 5190-5197

Research Article

5194

Fig.6. Real time plotting through Transmitter and Receiver modules and Mobile application

To record the mobile phone accelerometer data placed over the abdominal , a mobile application is used. Sensor Kinetics® Pro is an integrated display, recorder, and display for all of the main sensors on your Android smartphone. The software gives you a complete picture of the overall complexities of all the sensors' combined operations. Every sensor has a sophisticated Chart Viewer that helps you to record the sensor's data in great detail. The Multi-Sensor Recorder tracks multiple sensors at a controlled data rate at the same time. The software will store and exchange map data, as well as add specialised filters and configurations to push the limits of smart phone sensors. Using Bluetooth application data is sent to Arduino receiver. The Fig 6 portrays the accelerometer data collected from mobile phone as well as from developed DAQ system.

Fig.7. Real time plotting through Transmitter and Receiver modules and Mobile application

The data from the sensor (accelerometer) can be used to develop a technique for detecting anomalies in the data. Since the data is raw and real-time, the solution would be to use unsupervised learning concepts. The information available is unlabelled. The aim is to formulate a method for presenting the existence or irregularities in results. If the study reveals the presence of some anomaly, it specifies the health status of the individual whose data is being analysed. On the target data, the project employs the One-Class SVM technique. It's an unsupervised algorithm that deals better on data that hasn't been labelled. The judgement surface over the data is discovered using the one-class method.

Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 5190-5197

Research Article

5195

6. Results and Discussion

Using the embedded module using Arduino microcontroller to gather respiratory data from a 3 axis accelerometer and transfer the data to a local computer, as well as analysing the respiratory data collection using a Machine learning algorithm to recognise abnormalities. Creating a software application that can serve as a dashboard for physicians and users, allowing doctors to write prescriptions that will be shown in the user dashboard.

Fig.8. X Y and Z axis data of Normal person and Covid patient

Fig.8. Sensor Data of covid patient with and without oxygen support Oxygen

The designed hardware data is tested with the mobile phone recorded accelerometer data to validate the performance of the sensor as well as the monitored person. The statistical data revelas that the standard deviation in x axis data is more compared to other two axes.

Parameters Time X-Axis Y-Axis Z-Axis

Count 2400 2400 2400 2400

Mean 316.671531 -0.830663 -0.552021 9.577992

Std 182.849870 0.101630 0.292503 0.24253

Min 0.123000 -1.258150 -1.075000 9.421180

Max 633.133700 0.058660 0.453700 9.746790

Table.1. Statistical details of 3 axes accelerometer of Normal Person

Parameters Time X-Axis Y-Axis Z-Axis

Count 2400 2400 2400 2400

Mean 316.700158 1.474356 0.548002 9.458557

Std 182.873025 0.315931 0.088695 0.100220

Min 0.177200 0.053870 -0.05986 8.183380

Max 633.2795500 2.760510 1.139640 10.207680

Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 5190-5197

Research Article

5196

Fig.9. Scatter Plot and histogram anomalies two data sets

To analyse coughing results, the one-class SVM algorithm received the most attention. The Anaconda Navigator Distribution’s Python programming language is used to execute the algorithm. The SciKit Learn Python libraries are used in the data anomaly detection algorithm. The data will be pre-processed and scaled for further processing in the algorithm. The data is then fed into the algorithm. The data is classified by the gaussian kernel using a one-class SVM. The graphs depict plots of data with deviations marked by red dots on the X-axis, Y-axis, and Z-axis values. The correlation data obtained by comparison of two data sets one is the covid patient with oxygen support and without oxygen support are -0.231 and -0.241.

The scope of extending the work using the concept of telemedicine, to create an online application for remote consultation between a doctor and a patient[4]. Each patient can build an account on the web online application to access their report and doctor's prescription. The information administrator transfers the documented report to the net application, and the doctor then uploads the drug to the desired patient. Hence the real time data monitoring help to doctor to diagnose patient prior to reach the worse condition[8][9].

7. Conclusion

In the case of the COVID-19 pandemic, precise remote RR control is likely to be crucial. It may make healthcare assistance more accessible to self-isolated COVID-19 patients, as well as all patients who have limited access to medical services during this crisis. The introduction of prompt and cost-effective healthcare programmes, such as early detection of patient worsening, remote triage, and home surveillance of COVID-19 patients discharged from hospitals, will benefit from improved remote patient monitoring. This would minimise the pressure on hospitals, reduce the risk of infection among healthcare workers, and therefore viral transmission. The existence of a large quantity of reliable RR data would also aid in the creation of predictive models for hospital admission risk, as well as diagnostic and prognostic models. Finally, the work conclude that precise remote RR monitoring is critical and should be done in conjunction with other vital signs monitoring. This target can be accomplished with ready-to-use technical solutions. Effective RR tracking will be a big help in dealing with the latest COVID-19 situation and any situations that could happen in the coming months or years.

References

1. Preejith S pi, Ahamed Jee Janil, Paresh Maniyarl, Jayaraj Josephl and Mohanasank Sivaprakasam “Accelerometer Based System for Continuous Respiratory Rate Monitoring” IEEE Instrumentation Measurement Society 2017

2. S. Fleming et al., "Normal ranges of heart rate and respiratory rate in children from birth to 18 years of age: a systematic review of observational studies," The Lancet, vol. 377, no. 9770, pp. 1011-1018, Mar. 2011.

3. M. R. Ambekar and S. Prabhu, "A Novel Algorithm to Obtain Respiratory Rate from the PPG Signal," International Journal of Computer Applications, vol. 126, no. 15,2015.

4. Andrew Bates, Martin Ling, Christian Gengy, Alice Turky and D. K. Arvind ”Accelerometer-based respiratory measurement during speech” 2011 International Conference on Body Sensor Networks 5. P.D. HUNG1,2, S. BONNET1, R. GUILLEMAUD1, E. CASTELLI2, P. T. N. YEN2 “Estimation of

respiratory waveform using an accelerometer” IEEE International Symposium on Biomedical Imaging June 2008.

Turkish Journal of Computer and Mathematics Education Vol.12 No.10 (2021), 5190-5197

Research Article

5197

6. Dusarlapudi, K., Narasimha Raju, K., "Embedded prototype of 3 DOF parallel manipulator for endoscope application using 3 axis MEMS accelerometer", Journal of Advanced Research in Dynamical and Control Systems, Vol 12, issue 2,1225-35, Mar 2020.

7. Uday Kiran, K., Shakeela, S., Kalyan, D., ...Sai Chaitanya, G., Ekanath Reddy, G.,Detection of abnormalities in covid-19 infected human respiratory system using accelerometer with the aid of machine learning, International Journal of Pharmaceutical Research, 2020, 12(4), pp. 3399–3407.

8. Preeti, M., Guha, K., Baishnab, K.L., Dusarlapudi, K., Narasimha Raju, K., Low frequency MEMS accelerometers in health monitoring – A review based on material and design aspects, Materials Today: Proceedings, 2019, 18, pp. 2152–2157.

9. Narayana, M.V., Dusarlapudi, K., Uday Kiran, K., Sakthi Kumar, B., IoT based real time neonate monitoring system using Arduino, Journal of Advanced Research in Dynamical and Control Systems, 2017, 9(Special issue 14), pp. 1764–1772.

10. Kalyan Dusarlapudi, K Narasimha Raju, Preeti M, ASCS Sastry, Genesis of MEMS Accelerometers for Select the Optimal Accelerometer for Bio Applications, International Journal of Innovative Technology and Exploring Engineering,2019, Volume-9 Issue-2, December 2019