The Design and Construction of a 12-Channel Electrocardiogram Device Developed on an ADS1293 Chip Platform

Abstract

An accurate and compact electrocardiogram (ECG) device will greatly assist doctors in diagnosing heart diseases. It will also help to address the increasing number of deaths caused by heart disease. Accordingly, the goal of the project is to design and construct an easy-to-use compact 12-lead electrocardiogram device that communicates with a computer to create a system that can continuously monitor heart rate and which can be connected to allied medical systems. The design is based on an ECG receiver circuit utilizing an IC ADS1293 and an Arduino Nano. The ADS1293 has built-in input Electromagnetic Interference (EMI) filters, quantizers, and digital filters, which help in reducing the size of the device. The software has been created using the C# programming language, with Windows Presentation Foundation (WPF), aiding the collection of the ECG signals from the receiving circuit via the computer port. An ECG Multiparameter Simulator has been used to calibrate the ECG device. Finally, a plan has been developed to connect the arrangement to health systems according to HL7 FHIR (Health Level Seven Fast Healthcare Interoperability Resources) through Representational State Transfer Application Programming Interface (Rest API). The ECG device, completed at the cost of U$169 excluding labor, allows for the signal of 12 leads of ECG signal to be obtained from 10 electrodes mounted on the body. The processed ECG data was written to a JSON file with a maximum recording time of up to three days, managed by a Structured Query Language Server (SQL) Server database. The software retrieves patient data from electrical medical records in accordance with HL7 FHIR standards. A compact and easy-to-use ECG device was successfully designed to record ECG signals. An in-house developed software was also completed to display and store the ECG signals.

1. Introduction

In Vietnam, wherein present studies have been carried out, cardiovascular disease accounted for more than 170,000 deaths in 2019 [1]. Globally, according to the World Health Organization (WHO), some 17.9 million people died in 2016 due to cardiovascular diseases (CVDs), about 31% of all deaths [2]. Diagnostics to detect heart disease is obviously a prime concern, there being many devices used to assist clinicians in making decisions. In particular, devices measuring electrical heart signals are indispensable in the process of diagnosing a range of heart diseases (including arrhythmia, heart failure, and coronary artery disease). Common to hospitals worldwide, ECG devices offer a range of advantages, including being simple to use and rapid in providing data, and devoid of side effects to patients and medical staff during the process of noninvasive measurement [3,4]. However, the current popular forms of ECG devices still have some limitations, including with respect to cost and size as well as weight. Moreover, there are additional limitations concerning the recording time of the signals and the ability to store data. Integrated circuits such as ADS1298 and ADS1293 have been developed to overcome such limitations, replacing traditional electrical circuits, reducing the size of devices, and allowing greater applicability in responding to potentially life-threatening heart conditions [5,6,7]. The introduction of reduced-cost high-performance microcontrollers should also be acknowledged, for example, Arduino, PIC, and STM32, all in all making it easier to design flexible ECG devices.

In general, the majority of ECG recording devices presently print results on paper to allow physicians to diagnose heart disease, making it difficult to store and manage the output. The software supporting the machine also has attendant difficulties in communicating with other systems. Accordingly, in seeking to support the study and diagnosis of cardiac diseases and heart-related diseases, there is a burgeoning need for a receiver that: (i) can collect from multiple cardiac leads; (ii) is easy to carry, and; (iii) can be connected to computer software to display and manage data. Currently, in seeking to solve the problem of size, it is possible to use the ADS1293, an Analog Front-End chip that is also an integrated circuit dedicated to receiving and digitizing ECG signals, being not only small in size but also flexible in application [8,9,10].

For desktop software, the Windows Presentation Foundation (WPF) is a powerful platform used to design interfaces, process data, and support diverse data exchange methods, while Health Level Seven Fast Healthcare Interoperability Resources (HL7 FHIR) is a resource management standard used by many countries for digital health systems as an alternative to the HL7 standard [11,12,13]. In adopting these, the goal of the project herein has been to design and construct a 12-lead ECG device to communicate with a computer using the above technologies. A system is desirable, from signal acquisition hardware through to software for both observation and storage, saving time and being simple to manage as well as a source of data for research [14].

2. Material and Method

The following sequential steps have been taken to achieve the stated goal:

Step 1: Design a block diagram and system hardware to include: a 12-lead heart signal receiver circuit with an ADS1293 Analog Front-End chip [15] to be connected to a computer via an Arduino Nano board integrated with an FT232L chip [16], with a power circuit for the ADS1293 from a Li-Po rechargeable battery, also with status display control circuits. The battery is the main power supply for all of the circuitry, avoiding powerline interference in the ECG signal. Moreover, it prevents patient electrical contact with the power supply to avoid hazards to the patient. Step 2: Build the hardware and program the Arduino for data receipt, data transmission and management tasks. Step 3: Apply the WPF platform written in the C# language to build tools on the computer, such as that required for processing, displaying, and recording signals. Data and information are to be managed using a SQL Server database. Set up is also allowed of a communication channel with an electronic medical record system in accordance with HL7 FHIR standards [17]. Step 4: Test the entire system by using the ECG simulator and by receiving real signals from volunteers.

2.1. Design Block Diagrams and System Hardware

The system consists of a number of components connected together, as shown in Figure 1, including:

• Signal acquisition block: this includes 10 electrodes and 3 analog front-end circuits (AFEC). The ADS1293 AFE is a specialized circuit for analog reception for use with amplifiers with noise-sensitive signals. It has built-in filters, digital interfaces, and integrated circuits dedicated to cardiac signal acquisition applications.
• Control block: this includes a heat sensor, an Arduino Nano microcontroller, and other components. These are responsible for receiving data from the signal acquisition block, performing signal processing to calculate heart rate, and sending it to the computer. Power supply management for the receiver block includes reading the voltage value of the battery, checking the temperature in the box, and switching off the power when the temperature approaches overheating.
• Power supply block: this includes a battery, a charging circuit, and a buck-boost circuit. The task is to supply power according to the voltage and current requirements for the signal acquisition block. In addition, the Li-Po battery will be recharged when the status shows a low battery, with the Arduino Nano microcontroller still connected to the computer.
• Computer block: this is provided as a personal computer running on the Microsoft Windows operating system. The role is to select the working mode for the control block, to receive the signals and to write these to a file that combines the receiver information and stores it in an SQL database known as the Electronic Health Record (EHR), according to the figure. Moreover, it is necessary to have an interface to interact with users, with C# chosen as the language for interface programming.

2.2. Hardware Construction and Arduino Programming

2.2.1. Hardware Construction

Using Solidwork software, a box and protective case have been designed for the circuit, including the charging port, USB port, reset button, power switch, and 4 status LEDs to show the power supply level, as shown in Figure 2.

The final product after assembling the box and cover is shown in Figure 3. The size of the box is 6.2 cm high, 8.5 cm wide, and 12 cm long. This box was created using a 3D printer.

2.2.2. Arduino Programming

Figure 4 is the main flowchart Arduino program that performs tasks when receiving commands from the computer. This includes reading data from the ADS1293, managing power to the ADS1293, and displaying the status. Since the data-ready pin of the ADS1293 connects to the external interrupt pin of the Arduino to ensure no data loss, an interrupt handler program is required that will read data from the ADS1293 and send the data to the computer. Three ADS1293s are connected to Arduino by using the SPI standard. Before powering the ADS1293, it is necessary to check the source and temperature. The MOSFET will need a 2 s delay to ensure a stable voltage and to perform the configuration. The loop (Figure 4) checks the corresponding commands sent from the computer.

Calculation of heart rate is by the method of the first derivative. The heart rate is calculated based on the Pan–Tompkins algorithm, which is used to detect the QRS complex in the ECG signal [18]. The operational diagram of the peak detection algorithm is shown in Figure 5, in which the Lead I of ECG collected signal is used for determining the heart rate.

2.3. WPF Platform Application Written in C# Language to Build Software on the Computer

The database is built as shown in Figure 6, including two parts for login management and information management. In the login management section, users will be allocated a username and password. Each username will have different rights in system operations. The “tbl_permision” is the table containing the permission group, including the following attributes: the authority group ID, the authority group name, and the authority group description to determine the rights of the user. Information management includes patient information, files including recording time, notes, ID number of the recorder, and other information. Every file that is logged will have a path written to the database for retrieval if needed.

2.4. Run a Full System Test

First, we use the designed hardware, receiving the signal from the SKX-2000 ECG signal simulator and observing the results on the computer. In addition, the SKX-2000 ECG signal simulator is an ECG Multiparameter Simulator that generates the standard ECG signal, and we use this signal to calibrate the ECG device. This standard ECG signal is measured and displayed on the monitor. Based on the ECG signal, we can check all of the ECG devices. Then, with the aid of a volunteer, an observation has been made of the signal received. The display is provided of just three leads, I, II and III, due to a limitation of the screen but nevertheless gives the necessary ECG information (Figure 7). However, one can switch the lead by selecting the other lead in the pop-up menu on the software to display another ECG signal lead on the monitor.

3. Results and Discussion

3.1. Results of Hardware Design and Construction

The detailed schematics of the acquisition circuit, control circuit, and power circuit can be found in Appendix A, Appendix B and Appendix C, respectively (Figure A1, Figure A2 and Figure A3). Additionally, the PCB designs for both the top and bottom view of these circuits are provided in Appendix D, Appendix E and Appendix F (Figure A4, Figure A5 and Figure A6). Figure 8 shows the completed three circuits used for building the ECG device. Each circuit corresponds to a specific function: acquisition circuit (a), control circuit (b), and power circuit (c).

Following the completion of the circuits shown in Figure 8, the assembled versions of the acquisition circuit, control circuit, and power circuit are presented in Appendix G (Figure A7). Subsequently, all the circuit components are carefully placed inside the outer casing, which is depicted in Appendix H (Figure A8). Upon completing the construction process, as in Figure 9, re-calculation was made of the entire cost of implementation, including components, labor, and the total cost for the project. The cost for the system hardware was 169 USD, excluding labor, accounting for some 80% of the expenditure for hardware. The hardware cost consists of three ADS1293, being the highest cost items, the Arduino Nano board, and other components.

3.2. Results of Building WPF-Based Application Software on Computers

The interfaces designed for the required functionality are shown in Figure 10 and Figure 11.

First, to use the ECG device, one must register on the system. Entering the correct username and password and pressing the “Log In” button, the user is then taken to the next interface.

3.3. The Results of Signal Reception on the Human Body

Figure 12 is a screen showing volunteer results from leads I, II, and III, processed with respect to the baseline and the low pass filter on C# software. As a result of the use of the software, a review of the recorded data is available within 40 s. The signal has been filtered for noise after passing the filters. The heart rate is also calculated with the value of 100, as shown in Figure 12.

This compact 12-lead ECG signal acquisition device has been completed at a cost-effective price. The device allows for long-term data collection and monitoring of ECG signals. Furthermore, the software system has been designed to display the waveform of the ECG signal, further allowing for the storage and retrieval of patient heart disease data. The system was tested on an ECG Multiparameter Simulator and collected actual data on volunteers. However, the system has not yet been used to measure actual patients under medical supervision. In the future, we aim to test on actual patients and receive clinician suggestions to improve the system.

The ECG prototype is designed to obtain ECG signals at a sampling rate of 200 Hz, using a 10-bit ADC and the ADS1293 input with a bandwidth of 40 Hz and a differential voltage range of ±400 mV at the input. The system is powered by a 2000 mAh Lithium Polymer battery, providing up to 12 h of operation time. Furthermore, the device is designed for easy recharging using an external 5V DC voltage source via the micro-USB charging port. Additionally, the system has a DS18B20 temperature sensor to provide for warning of overheating. When the temperature is over 50 °C, the buzzer is on. Furthermore, for any sudden power failure or other power issues, the system can be instantly reset via the reset push button. A switch is also designed to power off the device when not in use, saving the battery. The system does not have a mechanism to detect electrode disconnection. This needs to be observed through the signal displayed on the screen.

A low-pass filter with a cutoff frequency of 40 Hz is used to remove both powerline interference and high-frequency noise in the ECG signal. Specifically, an FIR filter with a Hamming window function and order of 161 is designed to remove high-frequency noise. Furthermore, the DC noise component is removed using a cascaded integrator–comb (CIC) filter (CIC as a moving average filter) with a delay of D = 5. By applying this filter, the obtained ECG signal contains the only ECG signal component with low noise and is displayed on the screen as well as stored in the database.

By default, due to screen limitation, the ECG signals that are displayed are from the three signal leads I, II, and III. Nevertheless, one can easily change the signal channels in order: for instance, to show the different leads of the collected ECG signal. The signals can also be selected for playback, as shown in Figure 13.

4. Conclusions

We have designed and built a 12-lead ECG signal collection device connected with WPF-based computer software written in C# to manage cardiac signals and patient information. This is a compact device of low cost and weight, providing for rapid display of results and offering the facility considerable ease in utilization. Further, testing is ongoing prior to seeking medical device certification. In future work, making use of the proposed ECG device within a remote heart disease diagnosis system is being considered. Specifically, the measured ECG signal will be sent to the server, use subsequently being made of a deep learning network heart disease classifier that has been designed to classify heart disease remotely.