Non-invasive Blood Oxygen Saturation Monitoring for Neonates Using Reflectance Pulse Oximeter

Transcription

1 Non-invasive Blood Oxygen Saturation Monitoring for Neonates Using Reflectance Pulse Oximeter Wei Chen 1, Idowu Ayoola 1 1. Department of Industrial Design, Eindhoven University of Technology, Den Dolech 2, 5612 AZ, Eindhoven, The Netherlands w.chen@tue.nl, i.b.i.ayoola@student.tue.nl Sidarto Bambang Oetomo 1,2, Loe Feijs 1 2. Neonatal Intensive Care Unit, Máxima Medical Center, 5500 MB, Veldhoven, The Netherlands s.bambangoetomo@mmc.nl, l.m.g.feijs@tue.nl Abstract Blood oxygen saturation is one of the key parameters for health monitoring of premature infants at the neonatal intensive care unit (NICU). In this paper, we propose and demonstrate a design of a wearable wireless blood saturation monitoring system. Reflectance pulse oxymeter based on Near Infrared Spectroscopy (NIRS) techniques are applied for enhancing the flexibility of measurements at different locations on the body of the neonates and the compatibility to be integrated into a non-invasive monitoring platform, such as a neonatal smart jacket. Prototypes with the reflectance sensors embedded in soft fabrics are built. The thickness of device is minimized to optimize comfort. To evaluate the performance of the prototype, experiments on the premature babies were carried out at NICU of Máxima Medical Centre (MMC) in Veldhoven, the Netherlands. The results show that the heart rate and SpO 2 measured by the proposed design are corresponding to the readings of the standard monitor. Keywords- neonatal monitoring; reflectance pulse oximeter; blood oxygen saturation monitoring; design process I. INTRODUCTION At a neonatal intensive care unit (NICU), continuous health monitoring for the neonates (i.e. new born infants) provides crucial parameters for urgent diagnoses so that adequate medical treatment can be instituted. Blood oxygen saturation is one of the key parameters to be monitored which assesses the percentage of arterial hemoglobine that is saturated with oxygen. Transmission and reflectance are two non-invasive techniques to perform pulse oximetry (SpO 2 ). Presently, at hospitals the oxygen saturation of the blood is monitored by a transmissive pulse oximeter with the sensor attached on the foot or palm of the neonate [1]. Placement of these sensors and the presence of all the wires lead to discomfort and even painful stimuli when the sticky sensors have to be removed [2]. Reflectance pulse oximeters attached on forehead [3, 4] have been developed. To optimize functionality and patient comfort, the design of a monitoring system is essential, which integrates all technical components into suitable material and forms, as well as moves sensors to an invisible background for better parent-children bonding. The Eindhoven University of Technology (TU/e) in the Netherlands has started a 10-year project on non-invasive perinatal monitoring in cooperation with the Máxima Medical Centre (MMC) in Veldhoven, the Netherlands. The goal of this research collaboration is to improve the healthcare of the pregnant woman and her child before, during and after delivery. A smart jacket integrated with textile sensors [5] and a power supply based on contactless energy transfer [6] have been developed in the project. In this paper, we present and demonstrate a design of a wireless blood saturation monitoring system for neonates at NICU using the reflectance pulse oxymeter based on Near Infrared Spectroscopy (NIRS) techniques. The reflectance method provides the flexibility for measurements at different locations on the body of the neonates, and has the potential to be integrated into a non-invasive monitoring platform, for example the smart jacket proposed in [5]. Prototypes are developed with the reflectance sensors embedded in soft fabrics and the thickness of device is minimized to optimize comfort. A standard oxymetry device - Avant 4000 Wireless Tabletop Pulse Oxymetry by Nonin was adapted for wireless transmission and advanced SpO2 processing. Experimental results from the testing on neonates at MMC show that the heart rate and SpO 2 measured by the proposed design are corresponding to the readings on the standard monitor. II. A. Design Process DESIGN PROCESS AND DESIGN CONCEPT Figure 1. Block diagram of the design process. The design process as illustrated in Fig. 1 begins with an orientation phase where clinical and design regulations are /DATE EDAA

2 closely studied. The results from this phase lead to an initial set of design requirements. In the later phase (i.e., C & D), two iterations are made. The process within each iteration encompasses concept generation, concept development, prototype building and testing. Knowledge and improvements from the first iteration are the basis of the second iteration and new requirements are re-established and implemented. B. Design Concept Figure 4. A - a strap to holder sensor components for testing purposes, B - Sensor components, C - extension cable to the wireless module, D - Sticker to firmly hold sensor for testing purposes. Figure 2. An illustration of the envisioned wireless monitoring during kangaroo mother care (KMC). The design concept is to integrate SpO 2 components into a non-invasive monitoring platform (e.g. a neonatal jacket) and wirelessly communicate to a monitor for display as illustrated in Fig. 2. Fig. 3 indicates the desired elements to be incorporated in the design, including the platform, interconnect architecture, hardware, software and the performance / functionality of the system. III. DESIGN DEVELOPMENT AND PROTOTYPE The prototype is intended to have a neutral form in order to be tested on all parts of the patient s body especially for validating the possible sensor positions. It consists of a set of components and a wireless module as seen in Fig. 5. The sensor head includes a red-infrared emitter and a photo diode integrated into textile (sensor head), an extension cable and a male plug. The wireless module is used as a local processor which sends signals wirelessly to the monitor (host system). It is also used to modulate and power the sensor components. Figure 5. Block diagram of prototype design. A. Measuring Technique Reflective SpO 2 Sensor There are two common techniques for SpO 2 measurement which are transmissive and reflective techniques. The reflective technique commonly known as Near Infrared Spectroscopy (NIRS) was chosen because it provides the flexibility for measurement at different locations on the neonates, and has the potential to be integrated into a non-invasive monitoring platform, such as the smart jacket [5]. While the transmissive technique only works sufficiently on thin body parts, such as earlobe, finger tip, neonate s palm and foot. Figure 3. General criteria intended to be integrated into the jacket. In the form giving aspects, the design is intended to have an adaptable style so as to enable tests on various body locations. This would help to determine appropriate sensor positions, placement techniques and best practices or methods for the development of the jacket. The diagram in Fig. 4 presents one of the many idea sketches for the prototype so as to elicitate critiques and testing for adequate development. Figure 6. NIRS technique the SpO 2 value is processed from the reflected light) [7]; (transmissive techniche shows light measured at the oppossite surface) [8].

3 As seen in Fig. 6, the key advantage of NIRS technique is the ability to sense the Red and Infrared light absorption on a single surface without having to place the sensors and emitters on opposite sides as in the case of transmissive techniques in Fig. 6. B. Wires and Electronics To build reliable electronics at the early stage for subsequent testing and validation is essential for this case design. Fig. 7 is a block diagram of the sensor module. It consists of a photo diode, Red & Infrared light emitting diodes (LEDs), an analog processing circuitry, Analog Digital Converter (ADC) for signal acquisition, an embedded microcontroller (µc), and an RF transceiver for wireless communication. The microcontroller could also have a built-in ADC. This unit should be small enough in order to integrate the entire module into the jacket. Meanwhile, the photo diode and Red & Infrared lights are integrated separately as the sensor head. A switch circuit is integrated in the sensor head for automatic switching when connected to the wireless module. Figure 7. Block diagram for the wireless module. Fig. 8 shows the connection circuit of the sensor head. The polarities of the red-infrared emitter are connected to the wireless module independently from the photo diode and a switch circuitry is embedded to activate the device when plugged. Figure 9. Block diagram of the receiver module (e.g. the monitor), which displays the final results. An Avant 4000 Wireless Tabletop Pulse Oxymetry by Nonin was adapted for wireless transmission and advanced SpO2 processing plus display in order to speed up the design process and guarantee the accuracy of results. C. The Layers and Materials As part of the functional aspects, the design is to satisfy safety, comfort, and reliability criteria. Topography for the layout of sensors, connectors, and wires was made to satisfy these criteria. Different layers were used as insulators, protectors or other functions. Fig. 10 illustrates the layer hierarchy, and assembled layers and components. In Fig. 10 layer hierarchy for sensor integration, layer 1 is the top layer which has direct contact with the skin. It is made of cotton with a few silicon spots which acts as an anti-slip to minimize false alarm caused by movement artifacts. Layer 2, is a transparent layer made of Nylon or transparent plastic which isn t reactive and doesn t deflect light. Layer 2 is used to protect the surface of the electrodes and prevents direct contact of the electrodes with the patient s skin. Layer 3 is a porous and semi elastic material which holds the electrodes and the connection circuitry unto the textile. Layer 4 holds the extension cable which connects to the wireless module while layer 5 serves as an insulator between layer 4 and 6. A switch is integrated to layer 6 for automatic switching when plugged and layer 7 insulates the layer from exposed conductors. Various connection and manufacturing techniques were explored to maximize the sensitivity of the electrodes. Figure 8. Connection circuit of sensor head, which includes a photo diode, Red & Infrared emitters and integrated switch. Fig. 9 presents a block diagram of the receiver module. It includes a microcontroller, an RF transceiver for dual communication and a data cable to be connected to the monitor unit where the advanced processing occurs. Figure 10. Layer hierarchy for integrating the device sensor into the neonatal jacket: 1. Top layer, 2. Protection layer, 3. Component Layer, 4. Extension layer, 5. Isolation layer, 6. Switch layer, 7. Base layer. Diagram of the assembled layers and components. D. Prototypes Two prototypes were made at instances of the two iterations realized. The first prototype was to affirm the electrical constructions were working fine and also to set an initial platform for further development. Fig. 11 is a picture of the

4 first prototype. It was made from pieces of soft fabric and textile wires with series of stacked layers amounting to a thickness of 2mm. Layer 1 and layer 2 as showed in Fig. 10 were not implemented here. Sewing techniques were used to create the circuit and as connectors between the components and the circuit without soldering. All materials used were flexible to enhance the comfort of the design. The LED acts as the light emitter while the photo diode was sensitive to only red and infrared lights and it performs the function of a light sensor, which detects the intensity of the reflected light and channels it to the local processor unit (e.g. the wireless module in Fig. 7) for further processing. Figure 13. Picture of the extension cable plus sensor head of the second prototype with the top layer made of soft cotton with silicon spots and the internal structure of the sensor head. More improvements of the second prototype over the first were made in constructing a safer, more reliable and efficient male header (plug or socket) to connect the sensor head to the wireless module (see Fig. 14). Figure 11. A photo of the first prototype. The second prototype (see Fig. 12 and 13) is the result of the second iteration which was to investigate better techniques for integrating the components into the jacket and to determine appropriate placement of the sensors. It was formed on the basis of the initial construction however different techniques were used. More efficient conductive, insulating and connective materials were used to increase robustness of design. For example, a 0.5 mm insulated copper wire was used to replace the textile wire; bootlace and soldering iron were used to ensure reliable connections; and a 0.1 mm flex wire was used to replace the thick extension cable. All layers were integrated into this prototype but layer 3 and 6 in Fig. 10 where merged together. A soft cotton material with silicon spots was used as the top layer which performs the function of an antislip to minimize movement of sensor. Moreover, it was able to be sterilized which is essential for safety. Figure 14. Connection header to wireless module of first prototype, connection header of second prototype. IV. EXPERIMENTS AND USER TESTS Validating and realizing specifications in this study is an important part of the process. Various experiments were carried out to test the functionality and performance of the system and some other tests focused on building the electronics for the prototypes. For the second prototype, three user-tests were conducted of which two were recorded. The first was on an adult to affirm the functionality, sensitivity and safety of the new system before being applied to neonates. The second and the third tests were to determine a suitable location for placing the sensors on the neonate and then, a more advanced integration could be achieved. A. Experiment Setup Prior to the user tests, we carried out sensor test to assure the correct performance of the integrated sensor head. The experimental setup for the sensor test (see Fig. 15) includes (i) a digital oscilloscope for data acquisition; (ii) a laptop for data reading; (iii) an electronic tool kit; (iv) an Avant 4000 Wireless Tabletop Pulse Oxymetry by Nonin for wireless transmission and advanced SpO2 processing plus display; (v) a function generator for modulating LED signal; (vi) photo diode, and (vii) red and infrared emitters. Figure 12. A photo of the second prototype placed on an incubator at MMC.

5 for a proper integration into the jacket. The focus was to determine positions in which the sensor could read the SpO 2 and heart rate signs accurately. The test was conducted at NICU of MMC Veldhoven. The device was tested on a premature infant, born after 30 weeks gestation in stable health condition. A neonatologist and an NICU nurse from MMC were present during the testing. With the permission of the hospital and the parents, the prototype was placed and observed on various locations of the neonate s body. Examples of the tested positions were: positions on the back, belly, chest, sides, legs, arm, neck and various locations on the head. Fig. 17 shows an instance of the test on a baby inside an incubator in the NICU at MMC Veldhoven. Figure 15. Setup for system development and sensor testing. The components for user test (see Fig. 16) include the following: (i) Video camera for recording the corresponding data from the standard device and the prototypes monitor; (ii) Digital camera for capturing instances during the test; (iii) Tripod stand for placing the video camera appropriately; (iv) Standard oxymetry device for validating the measurements from the prototype; (v) Prototype to be tested with monitor (vi) Test context. Figure 17. Photo of user test on baby inside an incubator in the NICU at MMC. Before implementing the test on the baby, the prototype was first tested on an adult and then sterilized to ensure that it was working properly. The results from this test are compared with the readings from the standard patient monitor. From Fig. 18 we can see that the meausrment of the corresponding signs were tightly synchronized therefore, conforming the accuracy of the prototype which gives a green light for testing on babies. Figure 16. Setup for user-test Participants for the user-test involved a medical doctor, a NICU nurse, a technical expert, a designer/researcher and test persons. This varied for each test according to the goals of different test. B. User Tests Tests were made on adults and babies in different cases. In this subsection, we report the test on adults and babies with the second prototype. For validating the testing results, an additional standard oxymetry device was employed and the SpO 2 and heart rate readings from the two devices were compared to verify the results of the test. The goal for conducting a user-test on babies in the incubator of NICU is to validate the function of the prototype and explore opportunities for placing the sensors on the body Figure 18. Recordings of the test on an adult for functionality and sensitivity of device. The test was conducted at the NICU in MMC.

6 After a pilot study we learned that the best results were obtained at the forehead, back of the head, back of the shoulders, neck, and foot which resulted in corresponding readings to the standard device. Examples of the test results on forehead and comparison with the readings from the standard patient monitor can be seen in Fig. 19. The consistent measurements on these positions were verified on another premature patient. These positions (e.g. forehead, back of head, back of shoulders, neck, and foot) are recommended for further research in placing the sensors, which thus opens design opportunities for an advanced integration. obtained. Position on the forehead, back of head, back of shoulders, neck, and foot the places of the sensors, that corresponded best with the readings on the standard SpO2 monitor. The results obtained from these tests indicate the need to improve the design in order to minimize false readings caused by movement artifacts. Other critical issues like power factor, advanced interaction, feedback control, tradeoff between performance and cost, etc. are circumvented, but there are still potential design opportunities for the future development of non-invasive neonatal monitoring. ACKNOWLEDGMENT The implementation of this research would not have been possible without the support, and endless efforts of a number of individuals. The authors are particularly grateful to Willem Jin as a co-team mate. Special thanks to Jeroen Veen and Martijn Schellekens from Philips Research Eindhoven for their instrumental support in facilitating this work both in material and technical solutions. We would like to take this opportunity to appreciate the great assistance of Geert van den Boomen of TU/e, and Astrid Osagiator together with the medical staff at Máxima Medical Centre, Veldhoven, the Netherlands for their valuable input and feedback to the project. Figure 19. Example of user test on a baby s forehead in the NICU at MMC. V. CONCLUSION At NICU, placement of skin adhesive sensors and the presence of wires lead to discomfort and even painful stimuli when the sticky sensors have to be removed. Thus, noninvasive neonatal monitoring is in demand for healthy development of the neonates. In this paper, we present our process and demonstrate a design of a wireless blood saturation monitoring system for neonates at NICU using the reflectance pulse oxymeter based on Near Infrared Spectroscopy (NIRS) techniques. Prototypes are developed with the reflectance sensors embedded in soft fabrics and suitable for the integration into a monitoring platform, for example, a neonatal smart jacket. Experiments were carried out on neonates at MMC and the results show that the heart rate and SpO2 measured by the prototype are corresponding to the readings on the standard monitor. To determine an optimal position for the sensors appeared to be a challenge. Series of tests were conducted to explore the effective positions and promising results were REFERENCES [1] I. Murković, M. D. Steinberg and B. Murković, Sensors in neonatal monitoring: Current practice and future trends, Technology and Health Care, IOS Press, vol. 11, 2003, pp [2] W. Chen, S. Bambang Oetomo, and L. M. G. Feijs, Neonatal Monitoring Current Practice and Future Trends, Handbook of Research on Developments in e-health and Telemedicine: Technological and Social Perspectives, IGI Global, to be published. [3] Y. Mendelson, R. J. Duckworth, and G. Comtois, A Wearable Reflectance Pulse Oximeter for Remote Physiological Monitoring, 28th Annual International Conference of the IEEE Engineering in Medicine and Biology Society (EMBS), Aug. 2006, pp [4] New OxyAlert TM NIRSensors, from Infant-Neonatal.asp. [5] S. Bouwstra, W. Chen, L. M. G. Feijs, and S. Bambang Oetomo, Smart jacket design for neonatal monitoring with wearable sensors, in Proc. Body Sensor Networks (BSN) 2009, Berkeley, USA, pp , [6] W. Chen, C. L. W. Sonntag, F. Boesten, S. Bambang Oetomo, and L. M. G. Feijs, A Design of Power Supply for Neonatal Monitoring with Wearable Sensors, Journal of Ambient Intelligence and Smart Environments-Special Issue on Wearable Sensors, vol.1, no. 2, , 2009, IOS press. [7] MedGadget, Internet journal of emerging medical technologies, from: onitor_approved.html. [8] D. J. Lynn-McHale, and K. K. Carlson (Eds.), AACN Procedure manual for Critical Care, Fourth Edition, W. B. Saunders Company, 2001.