Abstract: In this work, we describe the design and construction of a pulse oximeter for remote patient monitoring. The proposed device uses the ESP8266 microcontroller as the core component to establish the wireless connection and send information to a node.js server. Then, a web service shows the saturation of oxygen and the heart rate through a WebSocket connection. This paper describes the hardware and software designs, justifying the decision to select each component of the proposed device. Finally, some experimental results are shown to illustrate the performance of the developed prototype.
Keywords:PulsePulse, oximeter oximeter, WebSoket WebSoket, nodejs nodejs, microcontroller microcontroller, ESP8266 ESP8266.
Resumen: En este trabajo se describe el diseño y la construcción de un oxímetro de pulso para el monitoreo de pacientes a distancia. El dispositivo propuesto utiliza el microcontrolador ESP8266 como componente principal para establecer la conexión inalámbrica y enviar la información a un servidor node.js. Además, un sistema web puede mostrar la saturación de oxígeno y la frecuencia del latido cardiaco a través de una conexión con WebSockets. En el artículo se describen los diseños del hardware y software del dispositivo, y se justifica la selección de cada uno de sus componentes. Finalmente, se muestran algunos resultados experimentales que ilustran el funcionamiento del prototipo desarrollado.
Palabras clave: Pulso, oxímetro, WebSokets, nodejs, microcontrolador, ESP8266.
Resumo: Neste trabalho, descrevemos o desenho e a construção de um oxímetro de pulso para monitorar pacientes à distância. O dispositivo proposto usa o microcontrolador ESP8266 como o componente principal para estabelecer a conexão sem fio e enviar a informação para um servidor node.js. Além disso, um sistema web pode mostrar a saturação de oxigênio e a frequência cardíaca através de uma conexão com WebSockets. O artigo descreve os desenhos do hardware e do software do dispositivo e justifica a seleção de cada um dos seus componentes. Finalmente, são mostrados alguns resultados experimentais que ilustram o funcionamento do protótipo desenvolvido.
Palavras-chave: Pulso, oxímetro, WebSokets, nodejs, microcontrolador, ESP8266.
Original research
Pulse oximeter with Internet data visualization
Oxímetro de pulso con visualización de datos en Internet
Oxímetro de pulso com visualização de dados na Internet
Universidad ICESI
Received: 04 January 2018
Accepted: 02 February 2018
The pulse oximeter is a medical device used for the measurement of blood oxygen saturation SpO2 and heart rate. The measurement of oxygen in the blood has more than 150 years of history, with the advancement of technology, this device has had the opportunity to evolve, improve its accuracy and decrease its size, which has allowed access to a pulse oximeter to the population in general (López-Herranz, 2003).
The main foundation used in an oximeter is to use the physical properties of the reflection and absorption of light in a specific protein of the blood (hemoglobin), which when exposed to a certain wavelength, the way in which it absorbs and reflects the wavelengths of the light is different, depending on whether it transports oxygen [O2] or carbon dioxide [CO2]. From this variation, it is possible to calculate the percentage of oxygen in the blood.
The development of a pulse oximeter was proposed from the ESP8266 microcontroller, using the minimum electronic elements, to reduce the electrical consumption, reduce the size of the device and avoid the incursion of noise from analogous elements; the use of analogical components is replaced by digital ones, as is the case of filters. The device connects to a server wirelessly [WiFi] over the Internet and sends the signal captured by the sensor; subsequently, the data is filtered, processed and stored by a web application. The user can see these measurements stored or observe the measurement data in real time
Pulse oximeters are essential to move towards health monitoring devices remotely, either interacting with smartphones or integrated into the Internet.
Mejía Salas and Mejía Suárez (2012) present the design of a small and inexpensive pulse oximeter, suitable for portable applications, with real-time data transmission through a ZigBee wireless link or a USB cable connection to a host.
Gómez García and Velasco Medina (2014) describe the design of a pulse oximeter that incorporates configurable analog and digital blocks, which allow the adaptation of the signals supplied by the sensors and the digital signal processing to be carried out on the same chip. In addition, they present an Android app for the visualization and registration of biomedical signals in a local database, compatible with mobile devices with WiFi connectivity.
This paper describes the design of an oximeter that allows the transmission and visualization of data in real time in a web app.
Pulse oximetry is a measurement method used in different clinical areas such as surgical interventions; the control and treatment of COPD [Chronic Obstructive Pulmonary Disease]; the monitoring of asthma; the diagnosis and development of respiratory diseases, such as influenza; and in the intensive care units. The device uses the physical properties of light absorption of hemoglobin to determine the percentage of oxygen saturation in the blood (Li & Warren, 2012).
Hemoglobin is a protein present in the red blood cells, it is generated in the bone marrow and is responsible for transporting oxygen from the lungs to the tissues of the body, on its return to the lungs transports CO2, for its elimination (Peñuela, 2005). When hemoglobin carries oxygen, it is called oxyhemoglobin and it is an intense red color; when it transports carbon dioxide, it is called DE-oxyhemoglobin and it has a bluish red color.
The circulation of blood is due to the work of the heart, which pumps blood to every part of the body, so it can carry out its function of transport, protection, and regulation; hemoglobin is responsible for the transport of gases (Koolman & Röhm, 2005). The movement generated by the heart can be measured by counting the contractions that the left ventricle exerts in a certain time, this is called heart rate (Font, Pedret, Ramos, & Ortís, 2008).
To perform the measurement of oxygen levels in the blood, pulse oximetry uses two techniques: spectrophotometry, based on Lambert-Beer Law that uses the absorption of a certain light wave to determine the amount in a certain element, depending on of the structure of each molecule (Marczenko, Balcerzak, & Kloczko, 2000); and plethysmography, which measures changes in the volume of blood in the body. In the latter case, using optical methods is called photo-plethysmography (Abu-Rahma & Bandyk, 2012).
It is necessary to irradiate the two substances to be measured with two different wavelengths, in this case, the hemoglobin, which reacts differently when having oxygen or carbon dioxide present. As can be seen in Figure 1, oxyhemoglobin absorbs more light with frequencies of 910 to 940 nanometers, and DE-oxyhemoglobin does so in the red range of 640 to 660 nanometers.
Figure 1.
Absortion of ligth waves
The pulse oximeter can employ two types of detection: by reflectance, which is based on the light that reflects the blood; and by transmittance, which uses the absorption of light. In this project, the transmittance method is used, the device has two transmitters located in front of the sensor; the measurement is made on the finger (Webster, 1997). Figure 2 shows the structure of the device and the position of the components that comprise it.
Figure 2.
Diagram of the device clip
The Lambert-Beer law is the principle that uses the pulse oximeter to perform the measurement of SpO2 in the blood, this law expresses how the absorbance in a certain wave of light can be measured on a substance, which will depend on the concentration of the absorbent compound, the distance and the type of light wave.
The measurement is obtained by radiating one side of the finger with the corresponding light frequency in different time intervals; at the other end is the sensor that captures all changes in the signal.
The wave of light passes through the arterial blood, known as the dynamic or alternating part [AC] and the continuous or static part [DC] formed by tissues, bone, nail and skin, until reaching the sensor (Bencomo, Villazana, & Salas, 2016), as can be seen in Figure 3. We are interested in analyzing the AC signal, discarding the static component, this signal is only perceived when there is a pulse present, otherwise, it could not be detected, as mentioned.
Figure 3.
Components comprising the finger
Obtained the two signals we cannot directly use the Lambert-Beer law to use two wavelengths. Therefore, we use Equation 1, which determines the ratio that describes the difference between the two wavelengths:
Equation 1
The oximetry makes use of the radiation of two frequencies of light to detect oxygen saturation, using two light-emitting diodes, with the frequencies of 640nm-660nm (red) and 910nm-940nm (infrared). These LEDs turn on alternately in a defined time interval.
The sensor responsible for capturing the variation of light frequencies that pass through the user’s finger is the Texas Instruments OPT101 [TI]; the integrated one is a monolithic photo-diode that has a transimpedance amplifier that generates a reduced output in noise or peaks, with an effective wave frequency range ranging from 400nm to 1100nm (TI, 2015).
The obtaining of the signal of exit of the sensor for its sending and later processing must be converted to values that the device can manipulate, to the being transformed of its analogical form to digital it is needed that the converter is of high resolution, to differentiate the changes of the pulsatile wave, since only 2% corresponds to the values of the arterial blood of the signal obtained. The device used is the ADS1115, developed by TI. It has a 16 bit resolution, integrates a PGA for signal amplification, and carries communication with the microcontroller through the I2C protocol (TI, 2016).
The ESP8266 chip manufactured by Espressif originally started as a low-cost Wi-Fi module, capable of making TCP / IP connections, which was controlled by another microcontroller through AT commands. Later, the company released the SDK that allowed the chip autonomy to not depend on another microcontroller, the device could be programmed and operate autonomously (Espressif, 2017). Over time, they started using third-party firmware and the native GCC compiler allowed the device to run MicroPython, LUA, and Arduino.
The control of the LEDs and the capture of the photodiode is started. Figure 4 shows the signal used to control the emitting LEDs, each period takes 1 ms in which each state of high and low lasts 250 ns; to have a complete measurement and send the data is implemented in a cycle of 256 measurements.
Figure 4.
Control signal of the two emitting diodes
When the chip completes the reading, it proceeds to send the data to the server. The communication is carried out by the Web-Sockets protocol. This API provides bidirectional communication between client and server over a TCP connection (Chopra, 2015), the chip makes HTTP and TCP connections natively so that Web-Socket protocols can be implemented. To make a connection, initially the request is made in HTTP, in it, the client asks the server to update the connection. Its advantage is a continuous link to the server, which allows the transmission of data in real time, with little latency.
The physical development of the device was designed based on the technical specifications of the component manufacturers; the realization of the schematics and the printed circuits was carried out in the free tool Circuit-Maker of the company Altium Limited.
The chip originally has an ADC at 10bit of 1 volt, which is insufficient for use in the oximeter; the ADC 1115 is integrated in a single plate, together with the ESP8266, in addition to a current regulation system to work at 5 v. It was reduced to the minimum of components using only those necessary for the operation of the oximeter. Figure 5 shows the scheme of the proposed system.
Figure 5
Diagram of the pulse oximeter: connection of the ESP8266 chip with the ADC 1115
In Figure 6 you can see the PCB of the final device, 5 cm long and 3 cm wide. The device is powered by a micro USB connection, the choice of this input is due to the number of portable energy storage devices on the market.
Figure 6
PCB of the proposed chip for the pulse oximeter
For the implementation of the server and the web app of the oximeter, the platform of Node.js was used, it was decided to use this environment for ease of scalability and event handling. For communication through Web-Sockets, the server is responsible for receiving the signal captured by the microcontroller; once received, the signal is filtered and sent to the front-end for viewing.
A digital filter is a programmable system in charge of processing a signal previously converted to digital, an advantage of this type of filters is that it can be implemented in devices that have the necessary processing characteristics, leaving aside the use of analog devices.
Once the digital signal is obtained, a filtering process is necessary to eliminate frequencies outside the range used by the oximeter, this limit is established from the pulsations of the heart. The average heart rate per minute [ppm] for an adult at rest is between 50 and 100. People who exercise frequently can reach 30 ppm at rest, at most, the heart can support up to 220ppm, although this value varies depending on the age.
To maintain a relationship with respect to the maximum and minimum frequencies, he limited himself to measuring from 30 ppm to 300 ppm, which would correspond to 0.5 Hz and 5 Hz. For this purpose, an FIR bandpass filter of size 256 was designed. Figure 7 shows the response of the filter.
Figure 7
PCB of the proposed chip for the pulse oximetere
When the server receives the signal it filters and sends the two signals captured, the application processes the signal obtained by applying Equation 1, and obtains the ratio between the wavelengths; then the percentage is established through the generated curve, through a calibration process (in this case, the calibration was developed empirically, through measurements and comparisons).
As for the cardiac pulse, it would only be necessary to apply the self-correlation of the signal. The value of R is linked to the percentage of SpO2, once the equipment has been calibrated. This comparison data is stored in the microcontroller for its use.
Figure 8 shows the data generated by the proposed oximeter. The system uses libraries in JavaScript that were modified for the visualization of the photolithographic signal. Two canvases are used for oxygen levels and cardiac pulse.
Figure 8
Signal display
The construction of the device was carried out using the elements mentioned in the previous sections, the first phase of the measurement of the pulse oximeter is carried out by the sensor and the emitter; these were built from the scheme made.
Having the armed cards of the sensor and emitter, measurements can be made. These should be one in front of the other, for it, the two cards are assembled in a finger clip built with sheets of foamed PVC, a light, rigid and durable material. Figure 9 shows the finished device. This clip is 6 cm long, 2.5 cm wide and 3 cm high.
Figure 9
Device’s clip
The main core of the pulse oximeter formed by the ESP8266 device was placed on a wristband, just like the battery. In Figure 10 the final device is seen in operation.
Figure 10
Final device
The user can observe their oxygen levels in the blood and heart rate by entering the application site.
In this document was presented the prototype of a pulse oximeter based on the ESP8266 microcontroller with WiFi connection, which allows remote monitoring of oxygen saturation and cardiac pulse, it was possible to maintain a continuous monitoring, besides keeping a record of the same for further analysis. In the future, the aim is to improve the graphic environment, as well as the control of users, furthermore the implementation of this system to other vital signs.
Licenciado en Ingeniería en Ciencias de la Computación en la Benemérita Universidad Autónoma de Puebla (México), su principal área de interés en investigación es el desarrollo de elementos de Internet de las Cosas [IoT].
Licenciado en Ciencias de la Computación por la Benemérita Universidad Autónoma de Puebla [BUAP] (México, 2002); Maestro y Doctor en Ciencias de la Computación por el Instituto Nacional de Astrofísica, Óptica y Electrónica (México, 2003 y 2008, respectivamente). Desde 2008 es profesor-investigador en la Facultad de Ciencias de la Computación de la BUAP. Su investigación se enfoca en el desarrollo de modelos matemáticos y computacionales para el procesamiento de señales e imágenes con aplicaciones en biomedicina.
Licenciado en Matemáticas por la Benemérita Universidad Autónoma de Puebla (México); Maestro y Doctor en Ciencias Matemáticas por la Universidad Nacional Autónoma de México. Profesor-investigador de tiempo completo en la Facultad de Ciencias de la Computación de la Benemérita Universidad Autónoma de Puebla (México). Sus áreas de interés incluyen: combinatoria, teoría de grafos, computación cuántica y modelos matemáticos y computacionales para el procesamiento de imágenes.
Figure 1.
Absortion of ligth waves
Figure 2.
Diagram of the device clip
Figure 3.
Components comprising the finger
Equation 1
Figure 4.
Control signal of the two emitting diodes
Figure 5
Diagram of the pulse oximeter: connection of the ESP8266 chip with the ADC 1115
Figure 6
PCB of the proposed chip for the pulse oximeter
Figure 7
PCB of the proposed chip for the pulse oximetere
Figure 8
Signal display
Figure 9
Device’s clip
Figure 10
Final device
