Wearable, Multimodal, Biosignal Acquisition System for Potential Critical and Emergency Applications

For emergency or intensive-care units (ICUs), patients with unclear consciousness or unstable hemodynamics often require aggressive monitoring by multiple monitors. Complicated pipelines or lines increase the burden on patients and inconvenience for medical personnel. Currently, many commercial devices provide related functionalities. However, most devices measure only one biological signal, which can increase the budget for users and cause difficulty in remote integration. In this study, we develop a wearable device that integrates electrocardiography (ECG), electroencephalography (EEG), and blood oxygen machines for medical applications with the hope that it can be applied in the future. We develop an integrated multiple-biosignal recording system based on a modular design. The developed system monitors and records EEG, ECG, and peripheral oxygen saturation (SpO2) signals for health purposes simultaneously in a single setting. We use a logic level converter to connect the developed EEG module (BR8), ECG module, and SpO2 module to a microcontroller (Arduino). The modular data are then smoothly encoded and decoded through consistent overhead byte stuffing (COBS). This developed system has passed simulation tests and exhibited proper functioning of all modules and subsystems. In the future, the functionalities of the proposed system can be expanded with additional modules to support various emergency or ICU applications.


Introduction
e monitoring of biosignals is becoming increasingly important in modern healthcare and health management, especially for emergency or intensive-care units (ICUs) [1][2][3]. Patients with unclear consciousness or unstable hemodynamics often require aggressive monitoring by using multiple monitors. is type of monitoring is not only used to understand current physiological changes in the body but also provides a potential way to predict the occurrence of specific malfunctions (e.g., stroke or heart attack) [4]. Biosignals record a biological event (such as a beating heart or a contracting muscle) in both space and time. e electrical, chemical, and mechanical activities that occur during this biological event often produce signals that can be measured and analyzed [5]. Biosignals, therefore, contain useful information that can be used to understand the underlying physiological mechanisms of a specific biological event or system and may be useful for medical diagnosis [6,7]. ese signals can be acquired in a variety of ways [5,6,8]. Many developed products have combined different peripheral measurements (heart rate (HR), skin conductance level, etc.) into one actigraphy for healthcare [9,10]. However, very few studies or products have combined these peripheral activities with central biological changes, such as brain activities (electroencephalography (EEG), hemodynamics, etc.) [11][12][13].
Moreover, recent developments have shown that multiple biological signals can provide higher accuracy in terms of determining human performance [14,15] or diagnosing various diseases [16]. Biological signals in living beings are any signal (either electrical or nonelectrical) that can be continually measured and monitored to provide information about several different systems in the body [17]. For example, EEG can be used to assess brain function diseases, degenerative diseases, inflammation, epilepsy, and abnormal sleep [18][19][20][21][22].
In addition, electrocardiography (ECG) can be used to observe cardiac behavior and to detect arrhythmia, ventricular atrial hypertrophy, myocardial infarction, myocardial ischemia and other diseases [23]. It can reflect the degree of myocardial damage and the structure of the atria and ventricles [24,25]. Recently, oxygen saturation (SpO 2 ) combined with ECG was used to detect stenosis or occlusion of an arteriovenous fistula/graft, the progression of arteriosclerosis, and arrhythmia [25]. SpO 2 is a measure of the amount of oxygen-carrying hemoglobin in the blood relative to the amount of hemoglobin not carrying oxygen [26][27][28][29]. e body requires a certain level of oxygen in the blood to function efficiently [26,28,30]. In fact, very low levels of SpO 2 can result in severe symptoms [26]. An ECG conditioning circuit is employed to examine the measured ECG signal. A pulse oximeter with a finger-type probe is connected to indirectly measure the SpO 2 level, and then, the HR and SpO 2 level are obtained at the same time [31,32]. erefore, an urgent need exists to develop an integrated system that can record multiple biosignals simultaneously for medical applications. In this study, we develop a multibiological integrated monitoring system that can acquire and process multiple biosignals (such as EEG, ECG, SpO 2 , and HR signals) simultaneously. Additionally, the developed system can transmit integrated biological data wirelessly to a smartphone or PC and monitor EEG, ECG, SpO 2, and HR information in real time. Most conventional systems take data points separately and from different locations on the body. Additionally, complicated pipelines or lines increase the burden on patients and inconvenience for medical personnel. In contrast, we develop an intelligent sensor system that facilitates sensor fusion by sealing the electronics into a single mold. is design is less affected by motion artifacts. In the future, this option will provide flexibility to patients and give clinicians access to physiological data in real time. Moreover, the modular design concept (functional partitioning into discrete and scalable modules) will soon be applied to this system to expand the number of possible measurements and establish industry standards.

System Architecture.
e purpose of this study is to develop a biosignal recording system that can measure multiple biosignals synchronously (the system architecture is shown in Figure 1(a)). e system consists of three parts: a recording module, a microcontroller, and an online interface. e core system is the Arduino Mega 2560, which has the following characteristics: (1) an open-source single-chip microcontroller that uses an Atmel AVR single chip; (2) open-source software; (3) a hardware platform built into a simple output/input interface panel; and (4) the use of C language to develop the environment. e biosignals are acquired by separate related modules: (1) EEG module: BR8; (2) ECG module: AD8232; and (3) SpO 2 module: MAX30100.

EEG Signal Acquisition System.
e architecture of the EEG signal acquisition system is shown in Figure 1(b). e EEG signals are first amplified by two analog amplifier circuits and then sent to the analog-to-digital converter (ADC) (ADS1298) for sampling. e digital signals converted by the ADC are transmitted to the microcontroller unit (MSP430) via a serial peripheral interface (SPI). After the signals are encoded by MSP430, the EEG raw data are sent to other receiver devices through RS-232 for further analysis or application.

Front-End Active Circuit.
In the first stage, INA333 has a high input impedance (100 GΩ) and a high commonmode rejection ratio (CMRR) (100 dB). ese factors make INA333 suitable for the modular design of the integrated system. e transfer function of this design is as follows: . (1) In this circuit, R G � 14.7K is set as the gain of the amplifier. e preamplifier amplitude is set to 7.8 V/V, and the cutoff frequency is designed to be 0.103 Hz. us, the circuit achieves high performance in the first stage, as shown in Figure 2(a) (schematic of the instrument amplifier). In the second stage, OPA2333 has a low offset voltage (10 µV), a low quiescent current (17 μA), an excellent CMRR (130 dB), and a low power op amplifier. In this stage, OPA2333 is used as a bandpass filter. e second-order high-pass filter circuit is shown in Figure 2 e EEG signal is filtered by a second-order high-pass filter with a cutoff frequency of 0.108 Hz in the second stage. e gain of the high-pass filter is set to 174 V/V. e transfer function of this designed high-pass filter circuit is as follows:

Analog-to-Digital Converter (ADC).
In the ADC unit, ADS1298 integrates 8 channels of a high-resolution (24 bits) delta-sigma ADC and provides a high data rate (32 k samples per second). ADS1298 is usually used for ECG and EEG monitoring, as shown in Table 1.
Overall, the total gain of the active circuit is approximately 1,357 V/V, and the noise below 0.108 Hz is filtered out. e amplified and filtered EEG signals are digitized by the ADC unit (ADS1298). A total of 8 channel active circuits are operated at approximately 2.8 mA with a 3 V DC power supply. e details of the active circuits are shown in Table 1.

Microcontroller: MSP430.
After digitization is completed, the microcontroller unit (MSP430F5522, Texas Instruments, USA) receives the digital signals. MSP430 incorporates a 16-bit RISC CPU, peripherals, 10 KB SRAM, and 128 KB Flash. Dedicated embedded emulation logic resides on the device itself and is accessed via the Joint Test Action Group (JTAG) using no additional system resources. A sampling rate of 500 samples per second is set for the ADC.
MSP430 processes EEG signals and transmits them via a universal asynchronous receiver/transmitter (UART) interface to the Arduino board. Finally, the EEG data and other biosignals are displayed simultaneously on the screen and saved to the computer [8,33].

ECG Module.
AD8232 is a signal module for ECG measurement applications. It is designed to extract, amplify, and filter small biosignals under noisy conditions [34]. AD8232 is chosen because it has the best output impedance and gains. For the high-pass filter, a two-pole high-pass filter is used, and for the low-pass filter, a two-pole Sallen-Key low-pass filter is used [34]. AD8232 is available in a 4 mm × 4 mm, 20-lead LFCSP package. All the specifications make AD8232 suitable for use in the integrated system.

SpO 2 Module.
e SpO 2 module used, MAX30100, is an integrated pulse oximetry and HR monitoring sensor. It combines two LEDs, a photodetector, optimized optics, and low-noise analog signal processing to detect pulse oximetry and HR signals. MAX30100 operates from 1.8 V to 3.3 V power supplies [35] and can simplify the circuit design, reduce the system footprint, and reduce the design time and system power consumption; hence, it is suitable for the system [35].

Microcontroller Unit.
Arduino uses an open-source software and hardware platform and provides a simple I/O interface. It also supports development environments such as Java and C. In the experiment, we connected different modules to the Arduino Mega 2560. e Arduino board can accept input voltages from 7 to 12 V. In addition, it can output 3.3 V and 5 V to provide different physiological modules and has a sufficient number of I/O pins, making the integration of many physiological signal modules straightforward.
e Arduino Mega 2560 is an ATmega2560-based microcontroller board. It has 54 digital I/O pins (15 of which can be used as PWM outputs), 16 analog inputs, 4 UARTs (hardware serial ports), a 16 MHz crystal oscillator, a USB connection, and a power jack. It contains everything necessary to support the microcontroller and simply requires connection to a computer with a USB cable or boot-up with an AC-DC adapter or battery.

BR8 Connection to Arduino: Logic-Level Converter.
e Arduino board generally functions at 5 V, but the MSP430 used in our EEG system is a microcontroller operating at 3.3-3.7 V. If the TX/RX is connected, it will easily wear out or even overheat the MSP430 system (Figure 2(c)). Between the two systems, a logic-level converter is used for voltage conversion (Figure 2(d)).  e principle of this system is straightforward. When the current terminal (from the Arduino board) has a VL input of 5 V, the transistor does not conduct, and the back-end (to MSP430) VH output is 3.3 V. Conversely, when the frontend VL input is 0 V, the transistor is turned on, and the backend VH output is also 0 V; thus, the voltage can be converted to a level suitable for different systems. Combined with the logiclevel converter, the transmitting peak-to-peak voltage of MSP430 (Figure 3(a)) changes from 3.28 V to 4.8 V as it becomes the receiving peak-to-peak voltage of the Arduino board (Figure 3(b)). Additionally, the transmitting peak-to-peak voltage of the Arduino board changes from 4.8 V to 3.28 V as it becomes the receiving peak-to-peak voltage of MSP430.
Consistent overhead byte stuffing (COBS) is an algorithm for encoding digital packets to achieve efficient and reliable packet framing. It uses zero as a specific byte, which is inserted into a packet as a delimiter. When a zero data byte occurs, the operation replaces the zero bytes with a nonzero value. erefore, zero data are not present in the packet and are misinterpreted as a packet boundary (shown in Table 2).
is method enables the data packet receiver to quickly and efficiently restore the data packet to its original data form for efficient and error-free transmission. e data transmitted from MSP430 to the Arduino board are encoded by using COBS. To integrate different signals from each module, the packets need to be decoded into the original data. After the packets are decoded, the overhead byte and the delimiter byte are eliminated, and the zero value is reduced, as shown in Figure 3(c).

ECG Test Result.
In the ECG independent test, we use a UART to connect the AD8232 to the Arduino board through the TX/RX signal lines to receive the measured data from the experiments. e Arduino board then organizes the results into data for transmission to the computer as follows. e electrodes are grouped together to observe the activity of the heart from different angles. e typical ECG waveform can clearly be observed in Figure 4(a). PQRS and less obvious T waves can also be seen.

SpO 2 Test
Result. Unlike AD8232, MAX30100 uses I 2 C and the Arduino board to transfer data. rough the two signal lines SCL and SDA connected to the Arduino board, after the process starts, the red LED on MAX30100 is activated, indicating normal operation. e red LED plays an important role in measuring blood oxygen data. When measuring, we press the index finger over the red LED, allowing light to pass through the finger, with the reflected rays being monitored by the receiver. e signal  Emergency Medicine International received by the receiver is analyzed with respect to both the AC signal and the DC signal. In addition to measuring the blood oxygen concentration, MAX30100 can calculate the rhythm. Figure 4(b) shows the results of SpO2 transmission to the computer. e data in the figure are updated approximately every second. We carry out a simple test: the subject is first measured in a calm state, and then, the subject climbs the stairs to the eighth floor and is measured again. A simple comparison reveals that the heart rhythm is significantly improved after climbing the stairs.
e normal blood oxygen readings of an adult range from approximately 94% to 100%.

Graphical User Interface Test.
To monitor a variety of biomedical signals, we must develop an integrated interface that can simultaneously display EEG, ECG, and SpO 2 monitoring results. In this study, we use MATLAB to achieve our goal. MATLAB's powerful features provide us with a waveform interface and a collection of data packets from the COM port. e collected data are easy to store and use in subsequent calculations. In the experiment, we use the interface to display the EEG and ECG real-time waveforms, the blood oxygen level, and the HR. With MATLAB itself, we observe the individual waveform amplifications and store them at the end of measurement. e data for each experiment are available for later review.

Online Demo of the Integrating System.
In the integration interface test, the subject wears all the sensors of each module, including four EEG channels from BR8, three electrodes from AD8232, and one SpO 2 channel from MAX30100 ( Figure 5). Four EEG electrodes are placed on the forehead (Figures 5(a) and 5(b)), three ECG electrodes are placed on the chest (Figure 5(c)), and the subject's finger is placed on MAX30100 for the SpO2 measurement ( Figure 5(c)). When the preparation procedure is finished, the subject begins to undergo measurement of the EEG, ECG, and SpO2 signals simultaneously. e results we obtained are presented in the integrated interface shown in Figure 5(d).
e first to fourth channels are the signals received from the four electrodes of the EEG measurement. e fifth channel is the ECG signal, measured by the aforementioned three electrodes. e sixth and seventh channels are the HR and SpO 2 signals. In the experiment, we ask the subject to close their eyes for approximately 10 seconds and then quickly blink for approximately 10 seconds. As shown, for approximately 67-75 seconds, the EEG signal has a gentle waveform, which represents the brain wave of the closed eye. Between 75 seconds and 85 seconds, each peak represents an eye blink (Figure 6(a)). en, after enlarging the ECG signal, we observe that every complete cycle of the heart rhythm is a PQRST wave (Figure 6(b)). Finally, the HR and SpO2 values fall within the normal range ( Figure 6(c)).

Conclusions
is study developed a multiple-biosignal recording system based on a modular design that integrates EEG, ECG, and SpO 2 modules for medical applications. e data from these three modules are transmitted to the microcontroller unit, an Arduino board, through a designed logic level converter.
e Arduino board uses COBS to simultaneously decode the data and send them together to a customized graphical user interface for near-real-time presentation in a MATLAB environment, which is capable of receiving and visualizing data in real time and providing data for offline analysis. is system passed all simulation signal tests and functioned as intended when applied to a real human being. e modular design can endow the integrated system with high flexibility and enable our EEG module to be successfully combined with other commercial measurement devices on a common Arduino platform. In the future, we will replace our microcontroller with a more advanced microcontroller (e.g., the Arduino Zero) to increase the speed of data transmission. is improvement can help to enable the measurement of different biosignals (e.g., electromyography) or channels of the EEG module in this system for complete monitoring of human health status. Moreover, Wi-Fi/Bluetooth will support mobile platforms to increase the convenience of usage and application. We anticipate this multiple-biosignal acquisition system to be used for various real-life medical applications, such as in emergency or ICU settings, and to improve human quality of life.

Data Availability
Data will be provided on request through the corresponding author of this article.

Conflicts of Interest
e authors declare no conflicts of interest.