C4.2 Simplifying medical routine with wireless sensors Moor, Claudius Corscience GmbH & Co. KG Henkestr. 91, 91052 Erlangen Abstract Wireless near-field transmission has been a challenge for scientists developing medical sensors for a long time. Here, instruments which measure a patient s ECG, oxygen saturation, blood pressure, peak flow, weight, blood glucose etc. are to be equipped with suitable transmission technology. Application scenarios for these sensors can be found in all medical areas where cable connections are irritating for the doctor, patient and other care personnel. This problem is especially common in sport medicine, sleep medicine, emergency medicine and intensive care. Based on its beneficial properties with regard to power consumption, range, data security and network capability, the worldwide standard radio technology Bluetooth was selected to transmit measurements. Since digital data is sent to a receiving station via Bluetooth, the measurement pre-processing now takes place in the patient sensor itself, instead of being processed by the monitor. In this article, a Bluetooth- ECG, Bluetooth pulse oximeter, Bluetooth peak flow meter and Bluetooth event recorder will be introduced. On the one hand, systems can be realized with these devices, which allow patients to be monitored online (ECG, pulse oximeter). These devices can also be integrated in disease management programs (peak flow meter) and can be used to monitor high-risk patients in their home environment (chest strap event recorder). Introduction The current standard in medical measurement derivation requires the patient to be completely wired with cables. Often, every single electrode or every single sensor is connected to a monitor with a cable. Due to all these cable connections, the patient can t move freely, and the signals are riddled with artifacts due to the electrode cables being moved and electromagnetic interference. One practical example for the obstacles caused by wires is the derivation of a 12-channel ECG in ergometry. An ECG under stress is recorded while the patient is exercising on a stationary bicycle or treadmill. Due to the patient s movement, the wires which hang between the patient and the monitor are susceptible to interference/artifacts. In addition, the patient on the ergometer can t move freely due to the cables attached to the ECG monitor. More complex measurement derivation systems can be found in emergency medicine and intensive care. A multitude of sensors are used in parallel and are visualized centrally on one monitor. The standard instruments used include a 3-channel ECG, a pulse oximeter and a blood pressure instrument. It isn t hard to imagine that all these cable connections immobilize the patient and lead to measurement artifacts (see Figure 1). Figure 1: Classical signal derivation in intensive care The cable connections between the sensor and monitor are being replaced by wireless transmission technology. In the selection of this technology, extreme environmental conditions must be taken into consideration. The function of the sensors has to be guaranteed, even if an obese patient in a sleep lab, for example, is lying directly on the transmitter, thereby shielding it. In addition, the demands on the transmission security of the patient data must also be fulfilled (see Figure 2).
Figure 2: Wireless monitoring Materials and methods Structure of wireless measuring sensors: Current derivation systems consist of an electrode or sensor on the body, a cable and the measurement processing in the receiver (see Figure 3). The necessary cable connection can very well be several meters long. The non-amplified analog sensor signals along these meters of wire are susceptible to interference. Having reached the monitor, the analog signals, which are only in the lower millivolt range, are then amplified, filtered, and then digitalized by an analog/digital converter. Once they are in digital form, these measurements can be processed and displayed on the monitor. Standard signals, such as ECG, EEG, pulse oximeter parameters, and many more, are derived according to this principle. Figure 3: Block diagram of modern medical data transmission If one were now to design medical instruments for transmitting signals to a monitor using wireless transmission technology, the greatest challenge would involve the integration of the entire measurement pre-processing in the sensor. Analog amplifiers, as well as filters, ADCs, microcontrollers and a transmission module must be carried directly on the body (see Figure 4). Figure 4: Block diagram of wireless transmission Since the electronic components cannot be supplied with energy by ordinary cables, the only possible energy source which can be considered are batteries, either standard or rechargeable. To prevent the instrument from becoming too bulky, all components have to be highly integrated and designed to consume as little current as possible. Only this way can the instrument be comfortable to wear, which is the real advantage compared to the classical derivation systems. The power supply is either conventional AA/AAA batteries or rechargeable lithium batteries. Since the rechargeable lithium ion batteries have a high energy density (approx. 100 Wh/kg) compared to standard rechargeable batteries (NiMH approx. 24-40 Wh/kg) and are very small and available in many different shapes, they offer the best compact power supply [1]. They therefore also fulfill the demand for the compact form of instruments worn on the body. However, from case to case, it should be considered whether the higher cost of a system requiring Li ion rechargeable batteries is affordable for the patient. Since the measurement pre-processing now takes place within the sensor, intelligence can be integrated in the sensor in addition to the simple digitalization. It is then sufficient to send a calculated value over the wireless interface. The raw signal is only used for calculations in the sensor itself. This reduces the bit rate, which also leads to a reduction in current consumption. Selection of transmission technology: Because no visual connection can be guaranteed in medical instrumentation systems, only radio connections come into question for wireless transmission. It is also recommended to use a standardized form of transmission. This makes it easier to use already-existing radio modules, which already implement transmission protocols with error correction and are interference resistant. Also, standardized radio transmission has the advantage that different modules are compatible with each other. In this case,
it is sufficient for many applications to just equip the data transmitter, in our case the medical instrument, with a wireless interface. When personal computers, laptops or PDAs are used as a monitor, this standard interface is already integrated. In selecting the technology, it must also be considered that more than one instrument must be visualized on the monitor. This is made clear with the example of an emergency rescue, where the ECG, pulse oximeter parameters and blood pressure are among the signals to be derived. If one decides against a standard solution, preferring instead a proprietary solution, the above-described points must all be implemented oneself. One radio standard, which is already being used in hospitals, is Wireless Lan. Wireless Lan is used mainly for building up stationary wireless communication networks with high bandwidths. The range lies between 30 and 100 m in an office environment [2]. In clinics, therefore, laptops without an Ethernet interface access the local network over Wireless Lan. Due to the range and the high bandwidth, the power consumption is very high, which makes Wireless Lan out of the question for use in wireless medical sensors [3]. Zigbee technology will be of future interest, since it is currently an industrial standard in the USA (US standard IEEE 802.15.4). Zigbee is especially suitable for monitoring and controlling devices in home environments. A central transmitter can, for example, turn the lights on and off via remote control, operate a video recorder, or adjust the air conditioning. In industry, Zigbee is used in automation and control engineering applications [4]. Since Zigbee should implement a very low current consumption, it is conceivable that it could be used in medical technology. The standardization, however, is not so far advanced that modular solutions for initial evaluation are available. Bluetooth is used in a wide range of consumer products. This radio standard was originally implemented for communication between PCs, laptops, PDAs and mobile telephones and their peripheral devices. This also includes headset applications, which were developed with the goal of minimizing the power consumption. It is also necessary to have the option of making more than one connection to peripheral devices at the same time, for example to the mouse, keyboard and printer in the case of PCs. Bluetooth allows communication with up to 7 terminal devices [6]. It is also mandatory that the transmission to these products be error-free. Studies show that transmission errors occurring on the radio level are completely corrected, so that no further errors occur on the application level [5]. All three of the introduced radio standards have a justified use in medical engineering. Right now, however, Bluetooth is network-capable, has a low current consumption and is readily available. Therefore, Bluetooth fulfills the requirements, which are not fulfilled by the other standards. Bluetooth is the only technology qualified for use in wireless medical sensors at this time. To what extent Zigbee might also be able to be implemented in medical devices must be evaluated, as soon as appropriate modular solutions are available. Results BlueEKG Bluetooth ECG: BlueEKG is the mobile part of a Bluetooth ECG which can be used for the wireless measurement of 3, 6 or 12 ECG channels (see Figure 5). The introduced Bluesense module is used as the Bluetooth interface. The BlueEKG is designed for use in systems with online monitoring. In Variant 1, a 3-channel ECG is derived according to Einthoven by means of 4 electrodes. By means of a software configuration over the monitor, the 3-channel ECG can be extended to a 6-channel ECG. Variant 2 is a 12-channel ECG with 3 derivations according to Einthoven, 3 according to Goldberger and 6 according to Wilson. Besides the simple transmission of the ECG channels, the QRS complex is determined on the mobile part, the electrode contact quality and pace-maker pulse is measured, and. In addition, defibrillation protection is also integrated to protect against defibrillator pulses. During long-term monitoring, the measurement data is transmitted in real-time to a host system (e.g. PC, PDA, proprietary monitor). If the connection is for some reason interrupted during online transmission, the data is intermediately stored in the internal memory. Then it is attempted to reestablish the online transmission for 5 minutes, and to synchronize the intermediately stored data with the receiving system. This is the only way to guarantee complete monitoring, even when the patient accidentally moves out of the reception range. The heart rate and the current electrode contact of the individual electrodes are shown directly on the display of the device. In limited display mode, the heart rate is not displayed, but the heart symbol just blinks with the corresponding frequency. In addition to the optical display, if a QRS complex is detected, there is also an acoustical signal. If the BlueEKG is correspondingly configured, the instrument beeps with every R wave.
Figure 5: Bluetooth ECG BlueOxy Bluetooth pulse oximeter: The Bluetooth pulse oximeter BlueOxy is a compact pulse oximeter device, which can be worn on the arm and is network-independent (see Figure 6). It is used for the non-invasive measurement of oxygen saturation in arterial blood (SpO 2 ) and to measure the pulse rate. Besides the pulse rate and SpO 2, the measurement signal quality is shown in a bar graph on the integrated display. Online and offline modes are distinguished for long-term monitoring. In online mode, the measurement data is transmitted to a host system in real-time (e.g. PC, PDA, monitor) and is immediately evaluated and saved. As is done with BlueEKG, the data is buffered when the connection is broken. However it is attempted to reestablish the connection for 15 minutes instead of 5, and it attempts to transmit the intermediately stored data to the receiving system. In offline mode, the measurement data is saved in internal memory with the date, time and patient ID. The saved measurements can be evaluated offline and processed later using optional PC software. Furthermore, automatic measurements can be initiated in offline mode. Up to four wake-up times are configurable for measuring nightly values, for example. Figure 6: Bluetooth pulse oximeter BlueBELT Bluetooth event recorder for monitoring high-risk patients: The Bluetooth event recorder BlueBELT is a system for continuously monitoring high-risk cardiac patients in their familiar home environment (see Figure 7). In order to be able to get help quickly in the case of a critical heart state, the patient wears a chest strap, weighing only 120 g, which continuously analyzes the derived ECG. If the chest strap detects a life-threatening event, a Bluetooth connection is automatically established to a base station. This base station gives off an opto-acoustical alarm so that family members or care-givers are notified to rush to the scene. This guarantees quick assistance, which can save lives. In addition to alarming family members and care-givers, the base station also establishes a telephone connection to an electronic patient file. The file receives the ECG from the base station for the 10 s before and 10 s after the detected event. Also, the received ECG data received by the electronic file is also sent to a call center or rescue service. Thorough data analysis is carried out there by qualified personnel. For example, if ventricular fibrillation was detected, an emergency physician is notified immediately. Since BlueBELT is designed for long-term monitoring, the chest strap is not equipped with the standard adhesive electrodes, but with dry electrodes. These dry electrodes are made of V2A steel with a fractal coating to enlarge the surface area and to reduce the electrode impedance [7]. Compared to existing systems, BlueBELT also takes over the alarm in the case of critical cardiac states. This gives high-risk patients increased security when dealing with their heart disorders in their daily lives [8].
Figure 7: BlueBELT Bluetooth chest strap for high-risk cardiac patients Bluetooth in disease management, here an example with a Bluetooth peak flow meter: Application areas of wireless medical sensors in home monitoring can be found in sleep medicine, dialysis and the monitoring of high-risk cardiac patients. Furthermore, there are also application scenarios for wireless measurement registration in disease management programs (DMP) for CHF (congestive heart failure), high blood pressure, diabetes, as well as asthma and COPD (chronic obstructive pulmonary disease) patients. Disease management programs and home monitoring being carried out now use diaries, telephone interviews and data transmissions via modem or acoustic couplers for measurement registration. These programs require a high amount of personnel and perfect technical understanding on the part of the monitored patient. It is conceivable that incorrect data is transmitted due to a number of possible operating errors. If wireless sensors were used for measurements, the user wouldn t have to fill out forms, no cables would have to be laid between the instrument and the modem, and sensors wouldn t have to be pressed to the telephone receiver for acoustical transmission. The measurements can be transmitted wirelessly and automatically after measurement to an appropriately equipped modem without the patient having to do anything. This modem is installed one time and then the received values can be transmitted to the program operator. Depending on the application scenario, the measurement can be sent via an analog modem or GSM data connection or via SMS (see Figure 8). Figure 8: Concept for monitoring asthma and COPD patients The peak flow meter introduced here establishes a connection as a Bluetooth master to a mobile telephone after measurement and sends an SMS using GSM technology to the receiving server. Here, the peak flow is, the FEV1, as well as the medicine taken and any symptoms which occurred (coughing, phlegm, etc.) on that day are transmitted. This data is then automatically evaluated by the operator of the disease management program. If the measurements exceed or fall short of the set limits, the monitored person is contacted and, if necessary, is summoned to the clinic or sent to his general practitioner. This procedure carried out by the disease management programs allow deteriorations in the patient s health to be detected early. The patient consults a doctor in time, before the sickness gets worse. Summary and Outlook The sensors introduced here represent the state of the art of wireless Bluetooth-capable medical instruments. Right now, systems can be built up for use in online monitoring, monitoring of high-risk cardiac patients and to improve disease management programs. All of the introduced sensor systems minimize the immobility of the patient and his care personnel and they all reduce the artifact susceptibility. Furthermore, due to these new approaches for monitoring patients in their home environments, concepts can be realized which weren t possible in the past. The described obstacles and limitations are then history. In the areas of sport medicine, sleep medicine, emergency medicine and intensive care and home monitoring, the doctors and care giving personnel have more relief in their medical routines. An ordinary analog modem with a Bluetooth receiver will become established as a receiving station in the home environment. These bring a cost advantage compared to GSM data or SMS solutions. For mobile applications where the patient moves outside of a clearly defined living area, mobile telephones or smart phones can act as a suitable receiving station. From these receiving stations, the data is then passed on to a web server, which makes measurement evaluation with subsequent diagnosis possible. Medical
portals are realized this way, which makes integrated patient care possible through disease management programs. A further step involves the integration of as much intelligence as possible in the sensor itself. An example for this is the introduced pulse oximeter. It calculates the oxygen saturation from the measured optical transmission directly in the transmitter. It is therefore sufficient to simply transmit the final measurement. This reduces the bit rate over the wireless interface, which means that the electric power needed is also reduced. It must be carefully calculated, however, whether higher computer power can be achieved by a low-consumption microcontroller. A further advantage of integrated intelligence in sensors is the simple connection of instruments to different monitors. Since the monitors don t have to calculate the absolute values, the monitor acts as a pure visualization unit. This is especially helpful for the developers of OEM devices (OEM Original Equipment Manufacture) to differentiate their development and their knowledge. In addition, standard monitors are easy to modify, since only the Bluetooth interface has to be integrated to receive the data. Acknowledgement We would like to thank the Federal Ministry of Education and Research (BMBF) for supporting our development of wireless medical sensors in the cooperation project TEDIANET. Bibliography [1] K. Heinloth. Die Energiefrage: Bedarf und Potentiale, Nutzung, Risiken und Kosten. Verlag Vieweg, Braunschweig, Wiesbaden, 1997 [2] DAFU, Datenfunk in Deutschland (Wireless Lan), http://www.dafu.de, Homepage, Dezember 2004 [3] DPAC Technologies, Airborne 802.11b Serial Bridge Module, WNLB-SE-DP101, Datenblatt, Juli 2004. [4] A. Sikora, IEEE802.15.2 und ZigBee Drahtlose Low-Datarate-Mesh-Networks, Design&Elektronik, vol. 10/2004, pp. 126-133. [5] H. Pals, M. Gönne, H. Matz, H. Gehring, Untersuchung zur Störanfälligkeit von Bluetooth im perioperativen Bereich, Biomedizinische Technik, vol. 49-1, pp. 238 239, 2004. [6] J. Bray, C. F. Sturman, BLUETOOTH Connect Without Cables, Prentice-Hall, Inc., Upper Saddle River, New Jersey, USA. [7] A. Bolz, W. Urbaszek, Technik in der Kardiologie, Verlag Springer, Heidelberg etc, 2002 [8] M. Braecklein, C. Moor, I. Tchoudovski, A. Bolz, Erprobung eines automatischen Systems zur Kontinuierlichen häuslichen Überwachung von kardiologischen Risikopatienten, Biomedizinische Technik, vol. 49-1, pp. 234 235, 2004.