Conference Proceedings Second Joint EMBS-BMES Conference 2002 24thAnnual International Conference of the Engineering in Medicine and Biology Society Annual Fall Meeting of the Biomedical Engineering Society
Volume 3 Bioengineering - Integrative Methodologies, New Technologies
23-26October 2002 Houston, Texas, USA
Conference Co-chairs John W. Clark, Ph.D. Larry V. McIntire, Ph.D.
Program Co-chairs Periklis Y. Ktonas, Ph.D. Anthony G. Mikos, Ph.D.
WEEE
Publication Chair Fathi H. Ghorbel, Ph. D.
E&B -c-•
Copyright 0 2002 by The Institute of Electrical and Electronics Engineers, Inc. All Rights Reserved Copyright and Reprint Permissions: Abstracting is permitted with credit to the source. Libraries are permitted to photocopy beyond the limit of U.S. copyright law for private use of patrons those articles in this volume that carry a code at the bottom of the first page, provided the per-copy fee indicated in the code is paid through Copyright Clearance Center, 222 Rosewood Drive, Danvers, MA 01923.
For other copying, reprint or republications permission, write to IEEE Copyrights Manager, IEEE Operations Center, 445 Hoes Lane, Piscataway, New Jersey USA 08855-133 1. All IEEE Engineering in Medicine and Biology Society Conferences are peer-reviewed by a Technical Program Committee. The papers in this book comprise the proceedings of the meeting mentioned on the cover and title page. They reflect the authors’ opinions and, in the interests of time& dissemination, are published as submitted and without change. Their inclusion in this publication does not necessarily constitute endorsement by the editors, the IEEE Engineering in Medicine and Biology Society, or the Institute of Electrical and Electronics Engineers, Inc.
IEEE Catalog Number:
02CH37392
ISBN:
0-7803-76 12-9
ISSN:
1094-687X
Additional Copies of this Publication a re Available from:
IEEE Service Center 445 Hoes Lane P.O. Box 1331 Piscataway, NJ 08855-1331 Phone: (800) 678-IEEE (732) 981-1393 Fax: (732) 981-9667 E-mail:
[email protected] ~
ii
Table of Contents Theme 1. Biological Signal Processing and Systems Analysis Track 1.I : Nonlinear Dynamical Analysis of Biosignals: Fractals and Chaos
Mapping of Chaotic Patternsfor Localization of EEG Abnormalities Hudson, Donna L.; Cohen, Maurice E.
1
Self-Similarity in a Human Balancing Task Cabrera, Juan Luis; Milton, John G.
3
Central and Autonomic Regulation of Fetal Heart Rate: Nonlinear Analysis After Vibroacoustic Stimulation Magenes, Giovanni; Signorini, Maria G.; Arduini, Domenico
5
Towards Unbiased Reconstruction of Noisy Chaotic Dynamics Using a Robust Hypersurface Minimization Technique Lu, Sheng; Chon, Ki H.
7
Fractal Dyanmics of Body Motion during Walking in Poststroke Hemiplegic Patients Sekine M.; Akay M.; Tamura T.; Agner S. C.; Higashi Y.; Fujimoto T.
9
Changes in the Non-Linear Dynamics of Heart Rate Variability Due to Foot Based Vibration While in the Seated Position Villanueva, Jr., Azael; Madhavan, Guruprasad; McLeod, Kenneth J.
19
Nonlinear Modeling of Physiological Systems with Multiple Inputs Mitsis, Georgios D.; Marmarelis, Vasilis Z.
21
Nonlinear Advanced Methods f o r Biological Signal Analysis Cerutti, Sergio; Signorini, S. G.
23
Nonlinear Eeg Analysis in Epi1epsy:'fromBasic Research to Clinical Applications Lehnertz, Klaus
25
Incorporating Nonlinear Signal Analysis Results Hudson, Donna L; Cohen, Maurice E.
28
Track 1.2: Intelligent Analysis of Biosignals: NNs, AI, Fuzzy Logic, Knowledge-based Algorithms Session 1.2.1: Fuzzy Logic and Neural Networks for Biological Signal Analysis
Session 1.1.2: Nonlinear Dynamics of Biosignals I1 Investigating the Nonlinearity offMRI Activation Data Laird, Angela R.; Rogers, Baxter P.; Meyerand, M. Elizabeth 11
Finite-Time Growth Rate to Characterize Heart Rate Variability of Idiopathic Dilated Cardiomyopathy Patients Vallverdli, Montserrat; Tibaduisa, Oscar M.; Martinez, Antonio; Cinca, Joan; Bay& de Luna, Antonio; Caminal, Pere
17
Session 1.1.3: Mini-symposium: New Advances in Nonlinear Dynamic Analysis Methods
Session 1.1.1 :Nonlinear Dynamics of Biosignals I
Robust Segmentation of Switching Dynamics in Time Series Feng, Lei; Chon, Ki H.
Reduction of Electrocardiogram Interference from Diaphragmatic Electromyogram by Nonlinear Filtering Liang, Hualou; Wang, Hongbin; Lin, Zhiyue
13
15 iii
Hybrid Fuzzy Logic Committee Neural Networks for Classification in Medical Decision Support Systems Reddy, Narender P.; Rothschild, Bruce M.
30
Evaluation of Stress Reactivity and Recovery using Biosignals and Fuzzy Theory Sul, A.; Shin, J.; Lee, C.; Yoon, Y.; Principe, Jose
32
Similarity Classificationfor Bio-signals using Fuuy Table Look-up by Distance and Eigenvector Yoon, Y.; Shin, J.; Principe, Jose
34
Next Generation Decomposition of Multi-Channel EMG Signals Nawab, S. Hamid; Wotiz, Robert P.; Hochstein, Lorin M.; De Luca, Carlo J.
36
Personal Fitting Procedurefor Cycle Ergometer Workload Control by Artificial Neural Networks Kuyu, Tohru; Shibai, Keisuke; Hayashi, Yoichi; Tanaka, Kiyoji
38
Unconstrained Monitoring of Prevention of Wandering the Elderly Masuda, Yasushi; Yoshimura, Takumi; Nakajima, Kazuki; Nambu, Masayuki; Hyakawa, Tomihiro; Tamura, Toshiyo
Pan American Health OrganizationlWorM Health Organization (PAHO/WHO) Role on Technology Management in Latin America and the Caribbean Hernandez, Antonio
Track 8.3: Clinical Engineering Career
Portable Physical Activity Monitoring System for the Evaluation of Activity of the Aged in Daily Life Makikawa, Masaaki; Asajima, Shuzo; Shibuya, 1908 Koji; Tokue, Rinzo; Shinohara, Hiromi Distributed Intelligent Architecture f o r Falling Detection and Physical Activity Analysis in the Elderly Prado, Manuel; Reina-Tosina, Javier; Roa, Laura
Session 8.3. I: Best Clinical Engineering Practices I
-
1910
Track 8.2: Technology Policy for Health Care Systems Session 8.2. I: Medical Technology Assessment and Planning Health Care Technology Management - Canadian Northwest Tenitories 1912 Taylor, Kevin B. Health Care Technology Management Challenges in Mexico: Increase Clinical Engineering Services and develop Telemedicine Nationwide. Velazquez, Adriana
1915
A Framework for Health Equipment Planning, Incorporation, and Management Wang, Binseng; Welsh, Joseph P.
1916
Planar Flexible Electrodesf o r Use in Wound Sterilization Piggott, John M.; Berney, Helen; Clair, Jim; Hofm,ann, Marco; Stam, Frank; Morrissey, Anthony; Sheehan, Michelle M.
1929
The State of Clinical Engineering in China Zhou, Dan
1931
Hospital’s Medical Technology Evaluation Process 1935 Blair, Curtis A.; David, Yadin B. Implementation of a Novel Virtual Patient Record Architecture Berler, A.; Pavlopoulos, S . ; Karkalis, G.; Sakka, E.; Konnis, G.; Koutsouris, D.
1936
Acudes: Arquitecture for Intensive Care Units Decision Support Palma-Mendez, Jose T.; Marin, R.; Campos, M.; Carceles, A.
1938
Session 8.3.3: Device Safety and Medical Errors
1918
Health Care Technology Management: Challenges, Faced by Estonia in the Globalizing Economical Environment 1920 Aid, Siim
Clinical Engineering in the Tucumth Public Health System: 1990-2001 Evaluation Rotger, Viviana I.; Rocha, Luis; Olivera, Juan M.
1927
Activity Based Costing Applied to Clinical Engineering 1933 Rocha, Leticia S.; Bassani, JosC W. M.
Session 8.2.2: Medical Technology Acquisition and Maintenance
Healthcare Survey of Pakistan; A Clinical Engieering Perspective Ali, Hussain A.; Raza, Salman T.; Enderle, J.
Statutory and Voluntary Registration The Impact on ClinicaUBiomedicalEngineering in Ireland Grainger, Peter B.; Smith, Meabh; McGivney, John
Session 8.3.2: Best Clinical Engineering Practices I1
1914
Health Technology Policy Development and Implementation in South Africa Molai, Nonkonzo T .
Healthcare Technology Policy for Developing Countries Judd, Thomas M.
1926
1906
1922
1924 xlvi
A Wearable Computing Based System for the Prevention of Medical Errors Committed by Registered Nurses in the Intensive Care Unit Windyga, Piotr S.; Wink, Diane M.
1940
Micro-Process Based Management of Medical Equipment Maintenance Bassani, JosC W. M.; Rocha, Leticia S.; Liiders, Marcus L.; Bizinotto, Wilson J.
1942
Proceedings of the Second Joint EMBS/BMES Conference Houston, TX, USA • October 23-26, 2002
Distributed intelligent architecture for falling detection and physical activity analysis in the elderly M . P r a d o , J. R e i n a - T o s i n a , L. R o a Biomedical Engineering Group, University of Seville, Seville, Spain Abstract- A novel approach for the detection of falls, the analysis of body postures, mobility and metabolic energy expenditure of elderly people has been developed. It is based on a distributed intelligence architecture, supported by a wireless personal area network (WPAN) which allows a full 24-hour supervision of the user, both indoor and outdoor home. An open design methodology lets the addition of new sensors for the on-line monitorization of other biosignals. In this paper general guidelines and design issues are reported, with special emphasis on the Intelligent Accelerometer Unit 0AU), based on a four-axis accelerometer, the signals of which are transmitted to the W P A N server (PSE) for on-line processing. The availability of three axis in the median plane provides an inclination measurement with high sensibility. The IAU can be worn like a patch, fixed to the back, at the height of the sacrum. A prototype of the IAU is currently under validation phase, in order to optimize signal transmission between IAU and PSE.
Keywords- Falls, four-axis accelerometers, personal area networks, telehealthcare, intelligent sensors, signal processing, elderly. I. INTRODUCTION Falls in elderly people cause serious problems, with a demonstrated relation to morbidity and mortality within this population group. Besides, they represent an important burden for the public health system expenditure. Present-day technical advances in MEMS, communications and microcontrollers/DSPs, allow the design of low cost portable solutions for the monitorization of movement and on-line falling detection. During the last years many attempts to design portable devices offering falling detection have been proposed in the area of home telehealthcare. The initial results confirm the suitability of current technologies to give solutions to this problem. Different studies have shown the possibility of detecting transitions among postural states together with kinetic or metabolic parameters, from acceleration records acquired with a sensor located on the waist or breast, by applying different analysis techniques [1, 2]. However, most of the papers following this approach are based on the post-analysis of data recorded in a portable datalogger. The paper of Mathie et al. [3] shows one of the few systems designed for the on-line measurement of posture, mobility, gait and metabolic energy expenditure. This approach uses a triaxial piezoresistive accelerometer fixed to a belt, above the anterior superior iliac spine. However the design does not allow the monitorization of the user neither in the bathtub, taking a shower, nor during the time intervals when the individual is in the bedroom with the belt unfastened, and
Contract grant sponsor: Instituto de Salud Carlos III (Spain). Contract grant number: FIS 01/0072-01
0-7803-7612-9/02/$17.00 © 2002 IEEE
furthermore, it is useless out of home. All these are situations with a high fall risk [4, 5]. In this paper we present the architecture of a system for falling detection and metabolic energy expenditure measurement, with capability to perform posture and movement analysis, designed for the full-day monitorization of individuals at home and outside home. II. METHODOLOGY According to the principle of operation there are three main classes of fall detectors. The simplest sensors are exclusively based on the inclination of the user. This way there is no need to introduce accelerometers or perform complex signal analysis. A representative example is the ambulatory fall monitor of T amura et al. [6]. The second class of devices adds an impact sensor (generally piezoelectric) as a previous stage to falling detection, followed by the detection of inclination. The smart falling sensor of Williams et al. [7] is an example of these devices, which have demonstrated a better reliability than the former class. The third type of fall detectors can be distinguished for performing a real-time analysis of static and dynamic accelerations measured on the body. These detectors may record kinetic and gait analysis during a normal activity. This capability allows research on the causes of falls, as well as detection of modifications in the daily activity of the individual related to a deterioration of his/her health state. The developed falling detection system belongs to the third class discussed above. Its aim is to obtain on-line posture, kinetic and physical activity parameters, and detection of falls, in a similar fashion to the approach of Mathie et al. [3]. However, we have included a full 24-hour monitorization capacity as an additional requirement, at home and outside home. Up to our knowledge none of the actual systems fulfill this requirement. To achieve this goal our approach is based on a WPAN [8]. This way it is possible to move the sensor away from peripheral elements, like the emergency button. The sensor can be fixed to the skin by means of an impermeable patch, located at the height of the sacrum, as several recent studies recommend [2, 3]. Fig. 1 shows a simple diagram of the anatomical median plane with the accelerometer axis z (vertical), x (horizontal pertaining to the plane) and m (equidistant to x and z). The y axis is orthogonal to the median plane. From the architecture point of view, the access of the user to the P SE, is required to be as easy as possible. This unit includes peripheral modules like an emergency button, a display to show the state of the IAU and optic/acoustic signals that confirm the transmission of events/alarms to the tele-healthcare center. Connection is established by means of a Remote Access Unit (RAU) plugged to a PSTN, if the user is at his/her home, or by a WPAN additional client, with a serial link to a mobile phone. The autonomy of the IAU is about a couple of months, and requires minimum
1910
maintenance. The IAU detects any hardware failure, and in this case either informs periodically the PSE with a masterslave protocol or resets itself. If the failure impedes communication with the PSE, this unit warns about this event after a time-out of several minutes. With regard to the analyzed signal, several studies have shown the feasibility to perform posture and kinetic analysis in elderly people by acquiring accelerations in a 5 Hz bandwidth and an amplitude of + 2g, whenever the sensor is located near the body gravity center. Taking into consideration the intelligence distribution of the system, a sample frequency of 40 Hz per channel has been selected for the IAU. This signal is analyzed in a very simple way by a low cost microcontroller within the same IAU, to detect high frequency accelerations. As the IAU is fixed to the skin, the acceleration components due jolting and bouncing of the sensor are minimized, therefore high frequency components may indicate an impact with a high probability. The IAU four accelerometer channels are reduced subsequently to only three channels with a sample frequency of 10 Hz. Every second, these channels are transmitted to the PSE, where a digital signal processing is performed in a DSP. The minimum resolution of the IAU has been established in a 5 % of g, a value compatible to the noise level of the capacitive sensors used in the design and which offers enough precision for the determination of the parameters. Every sampled value is encoded in a byte word.
iiiiiiiiiiiiiiiiii i v Median Plane . . . . . . Fig. 1: Diagram of the anatomical median plane illustrating the sensor elements together with their axis. III. RESULTS A prototype for the IAU has been prepared, following the methodology explained above. In order to measure accelerations, two ADXL202E units of Analog Devices have been used. These modules are arranged as shown in Fig. 1, with a 3-volt supply. The four-channel duty cycle modulated (DCM) output of the ADXL202E devices are sampled every 20 ms by ports RBT-RB4, of the MicrochipTM PIC 16LC66 8bit CMOS microcontroller. Both ADXL202E devices are restarted every 18 ms by the microcontroller with a turn-on time of 1.9 ms. Signal acquisition for the 4 channels takes only 0.5 ms, because of an improvement in the decoding algorithm proposed by the manufacturer. This algorithm is based on the interruptions generated by the microcontroller whenever a rising or falling edge is detected in any of the 4 RBx ports. The micro controller stays most of the time in sleep mode, and is awakened every 18 ms by a WatchDog module. This way, mean power consumption of the
micro controller together with the accelerators, can be reduced below 0.35 mA. With this strategy, the IAU offers autonomy of nearly 2 months working 24 hours daily, with a 500 mAh Li battery. Its size is approx. 5 cm diameter and 5 mm thickness. IV. DISCUSSION AND CONCLUSION A novel system has been developed for falling detection and the analysis of posture and movement of elderly people. The main advantage of our proposal compared to other recently published devices is the capacity to perform daily on-line discreet monitorization during a full day, indoor and outdoor home. A prototype of the IAU is currently under validation phase in laboratory, in order to optimize signal transmission between the IAU and the PSE. Although several alternatives have been studied for wireless communications at 400 MHz and 2.5 GHz bands, power consumption of the RF transceivers is still high and further work is being done to achieve the maximum autonomy by researching on the most appropriate transmission system. Likewise the material that covers the patch is also to be improved. REFERENCES [1 ] B. Najafi, K. Aminian, F. Loew, Y. Blanc, and P. Robert, "An ambulatory system for physical activity monitoring in elderly," presented at 1st Annual International IEEE EMBS Special Topic Conference on Microtechnologies in Medicine and Biology., Lyon, France, 2000. [2] C. V. C. Bouten, K. T. M. Koekkoek, M. Verduin, R. Kodde, and J. D. Janssen, "A Triaxial Accelerometer and Portable Data Processing Unit for the Assessment of Daily Physical Activity," IEEE Transactions on Biomedical Engineering, vol. 44, pp. 136-47, 1997. [3] M. J. Mathie, J. Basilakis, and B. G. Celler, "A System for Monitoring Posture and Physical Activity Using Accelerometers," presented at 23rd Annual International Conference of the IEEE Engineering in Medicine and Biology Society., Istanbul, Turkey, 2001. [4] F. Padilla_Ruiz, A. Bueno_Cavanillas, C. Peinado_Alonso, M. Espigares_Garcia, and R. Galvez_Vargas, "Frequency, characteristics and consequences of falls in a cohort of institutionalized elderly patients," Atencion Primaria, vol. 21, pp. 437-42,445, 1998. [5] J. H. Downton and K. Andrews, "Prevalence, characteristics and factors associated with falls among the elderly living at home.," vol. 3, pp. 219-28, 1991. [6] T. Tamura, T. Yoshimura, F. Horiuchi, Y. Higashi, and T. Fujimoto, "An ambulatory fall monitor for the elderly," presented at Proceedings of the 22nd Annual International Conference of the IEEE, Chicago, Illinois, USA, 2000. [7] G. Williams, K. Doughty, K. Cameron, and D. A. Bradley, "A smart fall and activity monitor for telecare applications," presented at Proceedings of the 20nd Annual International Conference of the IEEE, Hong Kong, 1998. [8] E. Jovanov, J. Price, D. Raskovic, K. Kavi, T. Martin, and R. Adhami, "Wireles Personal Area Networks in Telemedical Environment," presented at Proceedings of the IEEE EMBS International Conference on Information Technology Applications in Biomedicine, 2000.
1911
