Integrated Systems for distraction free Vital Signs Measurement in Vehicles

Transcription

1 Cover story automotive electronics Integrated Systems for distraction free Vital Signs Measurement in Vehicles Mobile vital sign recording enables a variety of applications such as prevention emergency detection or comfort functions. An integration of vital sign acquisition systems is desirable as these functions should be available also in cars. This issue is addressed by the Department of Micro Technology and Medical Device Technol ogy at the Technical University of Munich and the BMW Group in the joint project Fit4Age. This article presents an evaluation of a car-embedded platform for vital signs acquisition which was carried out on the road. 34

2 authors Lorenzo T. D Angelo is Research Assistant at the Institute of Micro Technology and Medical Device Technology AgeTech Group of the Technical University of Munich (Germany). Prof. Dr. Tim C. Lüth is Director of the Department of Micro Technology and Medical Device Technology of the Technical University of Munich (Germany). Introduction In the modern world cars are a crucial means of transport. More and more people use them every day; hence a significant usage time accumulates over a lifetime. Novel assistance systems aim not only to increase this time s safety but also its quality. Not only in view of the aging society must the safety aspect be expanded in order to include the occupant s health. It is desirable to be able to use the vehicle for health prevention and to be able to safely recognise medical emergencies. The goal is not only the preservation of health but also of mobility one of the prerequisites for independence in older age. Modern vehicles come with a variety of sensors electronics for data processing and user interfaces. Today we find many systems for measuring vital signs in the household or in sports outside of vehicles. The integration of these measuring techniques and the use of the technology available in the cars offer therefore a great potential in the development of innovative assistance systems ranging from prevention to emergency detection and comfort. After a review of existing work in the field of vital parameter measurement integration in cars and a task description a concept for the in-car vital data acquisition and processing is described. The focus of the article is mainly on the description of the evaluation of the system on the road. State of the Art Systems for measuring blood alcohol concentration and fatigue are already commercially available. As these systems are not in the focus of this investigation they are not further discussed here. Simple and reliable non-invasive sensors are required to measure vital signs while driving. Only in this way parameters can be measured without the active collaboration of the person which is the case when another activity is being carried out. As for now most of the systems for measuring vital parameters in cars developed in research have been used to detect the emotional state of a driver and its influence on the driving state. Other studies have been carried out to determine factors which cause stress (increasing traffic increasing speeds) and how they can be reduced. The development presented in [1] is used to record the electrocardiogram (ECG). Determining the stress level is done based upon the heart rate variability as an indicator of the activity of the parasympathetic nervous system. The electrodes are on the steering wheel shift lever and on the left arm rest in order to measure the ECG as continuously as possible. The system was later extended in [2] with the measurement of body impedance (electrodes on the steering wheel) and blood pressure (optical measuring system on the dashboard). The sensors connected to a PC where the evaluation is done via analog / digital (A / D) converters. In [3] ECG is used to detect stress. In the work dry electrodes on the steering wheel are compared with electrodes attached to the chest. The signals are sent to a PC via Bluetooth. The value used here as a reference for the driver s stress is the vehicle s speed. In another work a stress level determined using ECG electromyogram (EMG) skin resistance and respiration rate is compared to a reference value based on the driver s facial expressions using image processing [4]. A correlation between stress level and heart rate and skin resistance is found in most examined drivers. The system presented in [5] measured the driver s skin temperature blood pressure blood volume and differential skin temperature on cheeks nose and fingertips. Reference [6] suggests using physical and mental parameters to control the vehicle e.g. detecting when a driver is no longer able to control the vehicle and an emergency halt is initiated as a consequence. A vehicle seats integrated system for non-contact ECG measurement is presented in [7]. Another system is able to recorded ECG embedded in a car seat as well as other parameters on the steering wheel and is based mainly on the use of technical textiles [8]. ❶ gives an overview of the main features of the mentioned systems. In overall all the systems are able to acquire vital signs inside a car. Just a autotechreview September 2012 Volume 1 Issue 9 35

3 Cover story automotive electronics [1] [2] [3] [4] [5] [6] [7] [8] HR Variability CE CE TPP HR (Heart rate) CE CE CE TPP TPP CE CE ECG CE CE CE BE CE CE EEG EMG BE Body impedance CE Blood pressure OS OS BE Blood volume Skin resistance BE BE CE Respiratory frequency BHS BA Skin temperature BT CT O 2 saturation TPP RPP Communication W W WL W W W WL WL Data processing PC PC PC PC PC PC MCU MCU ❶ Sensors and devices employed in the state of the art (Legend: CE = Electrodes in car; BE = Electrodes on body; BHS = Hall sensor on body; BA = Anemometer on body; BT = Thermistor on body; CT = Thermistor in car; OS = Optical system; TPP = Transmissive Photoplethysmography; RPP = Reflective Photoplethysmography; WL = Wireless; W = Wired; PC = Personal computer; MCU = Micro Controller Unit) OS few of the referenced articles mention the early detection of health changes as a potential application. Often a large number of sensors are employed and partly connected directed to the driver s body which renders the system unsuitable for everyday use as it could interfere with the driving task. Most systems also need a PC for data analysis the sensors must be connected directly with and do not offer a module collecting the data and offering a common interface to the vehicle information system for data transferring and further processing. None of the presented system designed for everyday use was tested on the road. Task The aim of the present work is to develop a system for in-car vital parameter acquisition. Mainly commercially available sensors should be used and be embedded in such a way that they do not distract the driver while driving. Data acquisition must be done on a separate module offering an interface for data exchange with the vehicle. The system should also be able to acquire heart rate information from a commercial chest strap and from other body worn sensors via radio e.g. blood pressure measurement [9]. Two system variants should be developed: one fully integrated into the vehicle and one externally applicable and thus portable. Concept The overall concept of the developed system will be presented briefly here. The vehicle integrated system variant was already presented individually in [10] while the portable system variant was presented for the first time in [11]. An examination done in a driving simulator of the BMW Group was also described in those articles. It was carried out to explore the basic possibility of embedding vital parameter sensors 100 % 90 % 80 % 70 % 60 % 50 % 40 % 30 % 20 % 10 % 0 % p OX on p OX off p EL on p EL off ❷ System components and interfaces ❸ Sensor usage time portion during drive 36

4 Not at all Rather not Rather strongly Very strongly The system disturbs me The system makes me insecure 9 % The system distracts me The system gives me more confidence The system is integrated well The system can increase road safety No Rather no Somewhat Yes 43 % 24 % 24 % 0 % 20 % 40 % 60 % 80 % 100 % ❹ Results of system acceptance interviews ❺ How motivated do you feel to take a measurement (interview result)? inside the car and to identify possible correlations between cognitive load and vital parameters. Also an evalu ation has been presented which was carried out test the functionality of the vehicleintegrated sensors. The evaluation presented in this article refers only to the vehicle-integrated variant. A more detailed system description is given in the referred publications. 4.1 System Description Both in the integrated and in the portable variant the system can be broken down into three components: a sensor unit a receiver unit and the system component inside the vehicle information system. ❷ gives an overview of the available system components. The sensor unit depending on the variant is attached externally to the steering wheel or built inside it. It measures the vital signs (heart rate oxygen saturation and skin resistance) by contact and sends it by radio to the receiver unit. The receiver unit is a module receiving the measured data from the sensor unit. The portable variant comes with its own display screen and battery. The integrated variant is embedded in the car and connected directly to the vehicle information system. The measurement values are sent to the vehicle upon request. Both the sensor unit and the receiving unit are interchangeable: an external sensor unit can be used with an integrated receiver unit. Therefore hybrid systems are possible. The system component inside the vehicle information system consists of an interface for data exchange with the receiver unit and of a software extension of the vehicle information extending it by a new menu item. It allows communication between the system and the vehicle or the driver respectively. The user can view the recorded parameters on the information screen. A more complex representation than on the portable receiver unit is possible due to the screen size. Thus not only the current readings but also their course over time can be displayed. Evaluation The system described was tested on the road in the time frame October 26 th 2010 until November 11 th 2010 in the vehicleintegrated variant. For this purpose it was built into a test vehicle of the BMW Group. N=21 subjects (5 female 16 male) with an average age of 65 took part. Each test drive lasted about two hours per subject. 5.1 Motivation The main purpose of the experiment was to determine whether the system motivates the driver to measure vital signs and what proportion of the travel time is actually available as measurements depending on the p considered type of OX = sensor. In addition EQ. 1 a subjective evaluation of the system p EL = has been done conducting a survey on the test subjects. 5.2 Materials and Methods The experiment was carried out using a test vehicle (730d) of BMW Group which the integrated system variant was embedded into. A switch enabling to change between the original and extended vehicle information system was also built into the vehicle. After a safety briefing and a familiarisation to the test vehicle in standing and during a 10 minute drive each subject was asked to drive a pre-defined route (highway state road and urban area length about 16 km) 3 times. During the first two trips the vehicle information system was in turn switched either into the original or into the extended state (with display of vital signs). In these trips all subjects were asked to drive as they were accustomed to and to touch the sensors on the steering wheel at their free will. During the third trip a reference measurement systems was connected to the volunteers and they were asked to touch the sensors as often as possible. After all three trips the investigator asked the subjects about their system rating using a questionnaire. The time portions p OX and p EL of the measuring times at which valid values were available are calculated as follows: Σ N G (k) 1 OX(k) Ν k=1 OX N G OX (k) = { 0 OX(k) Ν Σ N G (k) 1 EL(k) Ν k=1 EL N G EL (k) = { 0 EL(k) Ν autotechreview September 2012 Volume 1 Issue 9 37

5 Cover story automotive electronics No data storage Emergency halt On removable storage device Wireless transmission to 3 rd party Emergency call Wireless transmission to driver 0 % 20 % 40 % 60 % 80 % 100 % ❻ What type of data export is preferred (interview result multiple answers)? 0 % 20 % 40 % 60 % 80 % 100 % ❼ Which intervention is preferred on detection of a medical emergency (interview result multiple answers)? where OX(k) and EL(k) are the outputs of the punctiform pulse oximetry sensor and of the electrodes on the driving wheel s circumference respectively. The measurement point in time is k and N is the total number of measurements. Those are defined as not being a natural number if no valid value is available. Only the first and second trips were considered when calculating these figures. 5.3 Results The results of the determining the portion of available measurement time are shown in ❸ broken down depending on the sensor type and vehicle information system (original or with integrated vital signs display). In overall the punctiform sensor supplied readings during 44 % of the driving time while the sensor on the steering wheel circumference supplied readings during 81 % of the driving time. These values didn t decrease significantly using the original vehicle information system without vital signal display. In the following the main findings from the interviews are listed: Most participants valued the system as not being disruptive unsettling or distracting ❹. More than half of the subjects felt motivated rather strongly or very strongly to take a measurement by the system ❺. More than 80 % of the subjects prefer exporting the data using a removable storage device ❻. More than 90 % of the subjects desire an emergency halting function if it is able to securely detect a medical emergency and to safely bring the vehicle to a halt ❼. Summary The presented system for in-car vital signs measurement was tested on the road with 21 users with an average age of 65. It was shown that in normal driving situations a punctiform touch-sensitive sensor supplied measurements during 44 % of the driving time while a touch-sensitive sensor mounted on the steering wheel s circumference supplied measurements during 81 % of the driving time. Most subjects considered the system as not disturbing or distracting. They felt motivated by the system to take measurements during the journey. The resulting demonstration system can be convicted into a product together with car suppliers and enable novel driver assistance features. References [1] Jeong I.C.; Lee D.H.; Park S.W.; Ko J.I.; Yoon H.R.: Automobile driver s stress index provision system that utilizes electrocardiogram. In: Intelligent vehicles Symposium 2007 IEEE pp [2] Jeong I.; Jun S.; Lee S.; Yoon H.: Development of Bio Signal Measurement System for Vehicles. In: Convergence Information Technology international Conference on pp [3] Lee H.B.; Choi J.M.; Kim J.S.; Kim Y.S.; Baek H.J. et al.: Nonintrusive Biosignal Measurement System in a Vehicle. In: Engineering in Medicine and Biology Society EMBS th Annual International Conference of the IEEE pp [4] Healey J.A.; Picard R.W.: Detecting stress during real-world driving tasks using physiological sensors. In: IEEE Transactions on Intelligent Transportation Systems Vol. 6 pp [5] Yamakoshi T.; Yamakoshi K.; Tanaka S.; Nogawa M.; Shibata M. et al.: A Pre-liminary Study on Driver s Stress Index Using a New Method Based on Differential Skin Temperature Measurement. In: Engineering in Medicine and Biology Society EMBS th Annual International Conference of the IEEE pp [6] Zocchi C.; Rovetta A.; Fanfulla F.: Physiological parameters variation during driving simulations. In: Advanced intelligent mechatronics 2007 IEEE/ASME international conference on pp [7] Leonhardt S.; Aleksandrowicz A.: Non-contact ECG monitoring for automotive application Medical Devices and Biosensors 5 th International Summer School and Symposium on pp [8] Heuer S.; Chamadiya B.; Gharbi A.; Kunze C.; Wagner M.: Unobtrusive in-vehicle biosignal instrumentation for advanced driver assistance and active safety. In: Biomedical Engineering and Sciences (IECBES) 2010 IEEE EMBS Conference on pp [9] D Angelo L.T.; Lohmann M.; Lueth T.C.: A new device for motion-aware ambulatory blood pressure measurement. In: Pervasive Computing Technol ogies for Healthcare Fifth International Conference on pp [10] D Angelo L.T.; Parlow J.; Spiessl W.; Hoch S.; Lueth T.C.: A System for Unobtrusive In Car Vital Parameter Acquisition and Processing. In: Pervasive Computing Technologies for Healthcare Fourth International Conference on pp [11] D Angelo L.T.; Parlow J.; Spiessl W.; Hoch S.; Lueth T.C.: Unobtrusive In-Car Vital Parameter Acquisition and Processing. In: Ambient Assisted Living Wichert and Eberhardt (Eds. 2011) Springer Berlin pp Read this article on

6