4th International Workshop on Wearable and Implantable Body Sensor Networks (BSN 2007)

The paper presents a planar compact antenna structure used in sensors aimed at healthcare applications. Antenna performance is numerically investigated with regards to impedance matching, radiation patterns, gain and efficiency. The compact size of the sensor causes the antenna to be susceptible to variable changes caused by the presence of lumped components. The study illustrated the importance of including full sensor details in determining and analysing the antenna performance. The body-worn sensor performance is also demonstrated and effects on antenna parameters are analysed, specifically radiated power, efficiency and front to back ratio of radiated energy. Radio propagation characterisation of the sensor operation in stand alone and on-body scenarios are introduced. Improvements are necessary in antenna design, matching circuitry and also sensor layout for better coverage and for obtaining maximum achievable communication range to produce efficient and reliable medical telemetry and monitoring systems.

This paper studies the energy efficiency and QoS performance of 802.15.4 and 802.11e MAC protocols for body sensor network applications. We simulated a stand-alone body sensor network, as well as co-existence scenarios where the body sensors operate in the presence of voice, video and IT traffic. Our results indicate that although 802.15.4 and 802.11e can provide an acceptable compromise between power consumption and QoS in some scenarios, there are situations (e.g. co-existence with video and heavy data traffic) in which both performance criteria can not be met simultaneously. This highlights the need for improving existing MAC protocols or designing new solutions that can provide both extremely low power and QoS for body sensor networks (BSNs).

Low power on-body communication is introduced. The human body surface is examined for the communication channel and 10KHz-100MHz frequency band, ‘bodywire’, is found to be effective in the wireless on-body communication. DCI is proposed to avoid any intentional ground electrode for the capacitive coupling. A CMOS transceiver chip for the on-body communication is fabricated and can achieve 2Mbps with 0.2mW power consumption. The architecture of the BSN controller is proposed and fabricated with CMOS. It has a 16b RISC and a schedule director with TCAM. It can separately control 254 sensor nodes and consumes 14uW in normal mode and 160uW in alert mode including leakage current. The fabricated chip is used to transmit MP3 data from the finger tip to the earphone to enjoy the music. In addition, the BSN controller can detect the emotion of the user by using the data from the sensor nodes transmitted through on-body communication channel.

Bioimpedance spectroscopy (BIS) enables the determination of the human body composition (e.g. fat content, water content). From this data, it is possible to draw conclusions about the person’s health condition. The measurement is carried out with at least four electrodes placed on the body. Nowadays, positioning and wiring of the electrodes can only been conducted by qualified personnel. Even the latest systems on the market are uncomfortable to wear and their use for mobile purposes is highly limited. The commercial BIS electrodes, not suitable for a long term use, may cause allergic reactions. Textile integration plays a role not just concerning the manufacturing of long term electrodes, but also concerning the integration of cables and other electrical components into a wearable and comfortable application. In this article, a portable BIS system combined with textile electrodes is presented as a possible future application. The first validation results of the portable BIS-device and textile electrodes are analysed and the suitability of application is discussed in order to develop a wearable bioimpedance spectroscopy system.

To obtain maximum unobtrusiveness with sensors for monitoring health parameters on the human body, two technical solutions are combined. First we propose contactless sensors for capacitive electromyography measurements. Secondly, the sensors are integrated into textile, so complete fusion with a wearable garment is enabled. We are presenting the first successful measurements with such sensors.

In recent years much progress has been made in the integration of physical transducers into clothing e.g. breathing rate, heart rate and temperature [1]. The integration of chemical sensing into textiles adds a new dimension to the field of smart clothing. Wearable chemical sensors may be used to provide valuable information about the wearer’s health, monitoring the wearer during their daily routine within their natural environment. In addition to physiological measurements chemical sensors may also be used to monitor the wearer’s surrounding environment, identifying safety concerns and detecting threats. Whether the clothes are looking into the wearer’s personal health status or looking out into the surroundings, chemical sensing calls for a novel approach to sensor and textile integration. In contrast to physical sensors, chemical sensors and biosensors depend on selective reactions happening at an active surface which must be directly exposed to a sample. Therefore issues of fluid handling, calibration and safety must be considered. This paper discusses the constraints in integrating chemical sensors into a textile substrate. Methods of fluid control using inherently conducting polymers (ICPs) are discussed and a pH textile sensor is presented. This sensor uses colorimetric techniques using LEDs controlled by a wireless platform. Some of the potential applications of wearable chemical sensors are discussed.

We demonstrate how modulated magnetic field generated by an simple LC oscilator can be used to measure joint angles, which are a key element in posture recognition and many motion analysis applications. Our method uses the same physical principle as large stationary motion trackers, however it applies the principle in a way suitable for a small, low power wearable system. It has the potential to be more accurate while being smaller and cheaper then inertial tracking (MARG) approaches that today are state of the art in wearable motion tracking. The paper describes the principle behind our method, discusses the advantages and problems and presents our prototype implementation. On a large data set with an hour of recording, hundreds of motions and tens of thousands of measurement-points we demonstrate, that even with an initial crude system implementation reasonable accuracy (between 6 and 9 percent average error) can be achieved.

This paper take the principle of wireless communication with body implants to the implementation and testing stage. This paper describes a module that can fit into an implant case and results of testing a wireless link with a body model.

When designing a body sensor network, the mobile phone is a natural design point for data aggregation services. However, there is a missing bridge that allows sensors to communicate with existing commercial phones, even if they already possess sufficient storage and processing capabilities. The PSI board is a small expansion module that interfaces body sensor networks with commercially available cell-phones through a standard MMC/SD slot. Adding this expansion capability allows researchers to extend phone devices with additional capabilities, allowing the devices to easily serve as the hub of a wearable body sensor network. The PSI board has an integrated switching mechanism allowing transparent access to an MMC/SD card, an embedded microcontroller, accelerometer, expansion connector, and an IEEE 802.15.4 radio that can connect to a variety of commonly available wireless sensors. The capabilities afforded by the PSI board enable several new applications that would be difficult to implement given the sensors and capabilities currently found in existing cell-phone platforms.

We present a framework for the automated generation of power-efficient state detection in wearable sensor nodes. The core of the framework is a decision tree classifier, which dynamically adjusts the activation and sampling rate of the sensors (termed groggy wakeup), such that only the data necessary to determine the system state is collected at any given time. This classifier can be tuned to trade-off accuracy and power in a structured fashion. Use of a sensor set which measures the phenomena of interest in multiple fashions and with various accuracies further improves the savings by increasing the possible choices for the above decision process.

An application based on a wearable gait monitor provides quantitative results. Comparing the decision tree classifier to a Support Vector Machine, it is shown that groggy wakeup allows the system to achieve the same detection accuracy for less average power. A simulation of real-time operation demonstrates that our multi-tiered system detects states as accurately as a single-trigger (binary) wakeup system, drawing substantially less power with only a negligible increase in latency.

A low power 16-bit RISC is proposed for body sensor network system. The RISC is designed of basic 3 stage pipeline architecture which has 28 instruction sets. Some special instructions are proposed for efficient applications. The lossless compression accelerator is embedded in the RISC to support the low energy data compression. The accelerator consists of 16×16-bit storage array which has vertical and horizontal access path. By using the accelerator the energy consumption of the lossless compression operation is reduced by 95%. The RISC is implemented by 1-poly 6-metal 0.18um CMOS technology with 16k gates. It operates at 4MHz and consumes 24.2uW at 0.6V supply voltage.

The popular Friis transmission formula is used to evaluate the amount of path loss in free space between the transmit and receive antennas for the design of wireless transceivers. It is could be also used to estimate the path loss for the link of the on-body network when there is no barrier and no body surface in between. Paying special attention to short-range communication, this paper develops composite expressions for path loss between two dipoles. Theoretical analysis is used to prepare an impedance-based model and an Sparameter-based model to extract the path loss. These expressions support the validity of Friis formula up to a certain separation distance, below which an error is induced, possibly resulting in miscalculation of required transmission power, sensitivity and margins, during the design of transceivers. Simulations and experiments were carried out to verify the proposed model, and their results match up well. The difference in values of path loss obtained by the proposed analytical model and Friis path loss formula in free space is highlighted. This is important to be aware of while choosing design parameters for transceivers that will be used in applications operating at short ranges as compared to wavelength at the frequency of operation.

Patient monitoring methods for the early diagnosis of one lung intubation (OLI) are non-specific and controversial. The aim of this study is to evaluate a new acoustic monitoring system for the detection of OLI. Lung sounds were collected from 24 adult surgical patients scheduled for routine surgical procedures. Four piezoelectric microphones attached to the patients’ back were used to sample lung sounds during induction to anesthesia and tube positioning. To achieve OLI, the endotracheal tube was inserted and advanced down the airway so that diminished or no breath sounds were heard on the left side of the chest. The tube was then withdrawn stepwise until equal breath sounds were heard. Fiberoptic bronchoscopy confirmed the tube’s final position. Acoustic analyses were preformed by a new algorithm which assumes a Multiple Input Multiple Output (MIMO) system, in which a multi-dimensional Auto-Regressive (AR) model relates the input (lungs) and the output (recorded sounds) and a classifier, based on a Generalized Likelihood Ratio Test (GLRT), indicates the number of ventilated lungs without retrieving the original lung sounds from the recorded samples. This algorithm achieved an OLI detection probability of 95.2% with a false alarm probability of 4.8%. Higher detection values can be achieved at the price of a higher incidence of false alarms.

The automatic step detection is a crucial component for the analysis of vegetative locomotor coordination during monitoring the patients with Parkinson’s disease. It is aimed to develop the algorithms for automatic step detection in the accelerometer signal, which will be integrated in sensor networks for neurological rehabilitation research. In this paper, three algorithms (Pan-Tompkins method, template matching method and peak detection based on combined dual-axial signals) are detailed described. Finally, these methods will be discussed by means of dis- and advantages.

Flexible microphone and earphone prototypes for wearable applications were developed by using ElectroMechanical Film (EMFiℳ)*. A suitable application for the developed headset can be, for instance, as accessory of rescue services or sport enthusiasts. Due its versatile properties, EMFi can be used both as microphone and earphone material. The sensor operation is based on thickness changes caused by an external force or pressure, generating charge and thus voltage on the electrodes. EMFi also works conversely, converting electrical energy to vibration and hence functioning as an actuator. In addition, measurement electronics for the microphone and earphone were implemented. Preliminary test measurements were carried out: the frequency response of the EMFi microphone was compared with the one of a reference B&K microphone. The EMFi microphone provides rather good response. Also subjective listening tests were done. For these measurements, the EMFi headset was integrated inside a neoprene hood used by the surface rescuers. Both with the microphone and earphone the quality of voice was sufficient. Based on the results, EMFi seems to be a promising material for some wearable audio applications.

Body sensor networks e.g., for health monitoring, consist of several low-power on-body wireless sensors, higher-level devices such as PDAs and possibly actuators such as drug delivery pumps. It is important that such networks can adapt autonomously to changing conditions such as failures, changes in context e.g., user activity, or changes in the clinical condition of patients. Potential reconfiguration actions include changing the monitoring thresholds on sensors, the analysis algorithms or the configuration of the network itself. This paper presents a policy-based approach for autonomous management of body-sensor networks using the concept of a Self- Managed Cell (SMC).

Ponder2

is an implementation of this approach that permits the specification and enforcement of policies that facilitate management and adaptation of the response to changing conditions. A Tiny Policy Interpreter has also been developed in order to provide programmable decision- making capability for BSN nodes.

Application specific design requirements for wireless sensor network have posed new challenges and need for revision of existing designs for its implementation in healthcare systems. These requirements are different from that needed for environmental, agricultural and industrial purposes. The proposed paper discusses the design issues in implementation of Query driven Healthcare monitoring system using wireless sensor network. Various MAC layers designs have been studied for their usefulness and compatibility with the requirements in the healthcare system. Topology and network layer design have also been discussed. Further the low and computational capabilities of wireless sensor nodes itself adds to the design complexity. Therefore some compromises have to be made in both the domains. We aimed to implement a healthcare system using wireless sensor node based on the above study. A hardware platform was designed for the use as wireless sensor node. In this system, a wireless sensor node attached on the human body provides ECG (Electrocardiogram) and body temperature from multiple patients in an ad-hoc network. A mote-based 3-lead ECG and body temperature monitoring system operates in wireless sensor network. Health data from multiple patients can be relayed wirelessly using multi-hop routing scheme to a base-station, following IEEE 802.15.4 standard for wireless communication. Unique id assigned to each mote is used to identify each patient in the network.

Medical examinations often extract localized symptoms rather than systemic observations and snap shots rather than continuous monitoring. Using these methodologies, one cannot discretely analyze how a patient’s lifestyle affects his/her physiological conditions and if additional symptoms occur under various stimuli. We present a minimally invasive implantable pressure sensing system that actively monitors long-term physiological changes in real-time. Specifically, we investigate pressure changes in the upper urinary tract per degree of obstruction. Our system integrates three components: a miniaturized sensor module, a lightweight embedded central processing unit with battery, and a PDA. Our tether-free system measures pressure continuously for forty-eight hours and actively transmits an outgoing signal from an implanted sensor node to a remote PDA twenty feet away. The software in this in-vivo system is remotely reconfigurable and can be updated when needed. Preliminary experimental results of the in-vivo pressure system demonstrate how it can wirelessly transmit pressure readings measuring 0 to 1 PSI with an accuracy of 0.02 PSI. The challenges in biocompatible packaging, transducer drift, power management, and in-vivo signal transmission are discussed. This research brings researchers a step closer to continuous, real-time systemic monitoring that will allow one to analyze the dynamic human physiology.

Vibrational energy scavenging using piezoelectric material is a viable method to provide sufficient energy for low-power wireless sensor networks. The applications for such devices in hospital settings as well as

in vivo

are abundant. Current devices are limited by both their design and material selection. This paper will address optimizing the design of microscale devices by showing how the device strains under input vibrations are directly proportional to its power output, and by proposing alternate designs which increase the strain distribution over more of the device volume. Finite element modeling (ANSYS®) was used to determine the strain distribution in a cantilever, modified cantilever, trapezoid, and spiral shaped piezoelectric microscale energy scavenging system. The increase in strain under uniform acceleration was determined to be 0, 29.2, 37.8, and 87.0%, respectively, over that of a simple cantilever.

This paper presents a system whose purpose is to monitor a patient continuously from indoor or outdoor environments. The system is based on a Bluetooth PAN, carried by the patient, whose central node, a smart phone, compiles information about patient’s location and health status. These data are encrypted to be sent to a server through Wifi or GPRS/UMTS. The system provides facilities to access to patient’s data, even from a smart phone by a J2ME application. It also allows to configure remotely the threshold values used to detect emergency situations.

Post surgical care is an important part of the surgical recovery process. With the introduction of minimally invasive surgery (MIS), the recovery time of patients has been shortened significantly. This has led to a shift of postoperative care from hospital to home environment. To prevent the occurrence of adverse events, the care of these patients is mainly relied on routine visits by home-care nurses. This type of episodic examination can only capture a snapshot of the overall recovery process, and many early signs of potential complication can go undetected. The development of Body Sensor Networks (BSNs) has enabled the use of miniaturised wireless sensors for continuous monitoring of postoperative patients. This paper examines the potential of processing-on-node algorithms for further reducing the wireless bandwidth, and therefore the overall power consumption of the sensors. The accuracy and robustness of the technique are demonstrated with lab experiments and a preliminary clinical case study.

Heart Variability Analysis (HRV) is not suitable for real-time processing on a resource-limited, single sensor network node, such as a Body Sensor Network (BSN) node, due to the high sampling rate (> 200

H z

) required to digitise ECG signals and the non-preemtable nature of operating systems such as tinyOS. Both reasons combined dictate that the processing of each sample needs to be completed withing the inter-sample period, typically 5 msec for ECG signals. This paper discusses a dual-layer real-time heart variability analysis algorithm. The top layer is invoked every time a sample arrives. This layer includes a real-time algorithm that delineates the significant part of the ECG signal, the QRS complex. The second layer, is near real-time and is invoked only when a potential QRS is detected, at a significantly lower rate that corresponds to the person heart rate. This layer is responsible for detecting R peaks, estimating the interval between two successive peaks and performs heart rate variability analysis in the frequency domain.

Our system outperforms traditional ECG processing algorithms because the top layer completes well within the 5 msec sample inter-arrival period, ensuring that no samples are lost. The bottom layer can be delegated either to an underlying background task or a second processor. Because it is invoked less frequently than the top layer, it results in a lower interrupt rate, allowing for more flexible processing.

This paper investigates the combined use of ambient and wearable sensing for inferring changes in patient behaviour patterns. It has been demonstrated that with the use of wearable and blob based ambient sensors, it is possible to develop an effective visualization framework allowing the observation of daily activities in a homecare environment. An effective behaviour modelling method based on Hidden Markov Models (HMMs) has been proposed for highlighting changes in activity patterns. This allows for the representation of sequences in a similarity space that can be used for clustering or data-exploration.

In this paper we discuss the possibility of performing a sleep evaluation from the heart rate variability (HRV) and respiratory signals. This is particularly useful for non standard sleep measurements based on wearable devices or special sensors inserted in the bed. The HRV and the respiration signals were analysed in the frequency domain and the statistics on the spectral and cross-spectral parameters put into evidence the possibility of a sleep evaluation on their basis instead of the classical polysomnography in a sleep laboratory. Additional information can be achieved from the number of microarousals recognized in correspondence of typical modifications in the HRV signal.

A new system for wireless monitoring of pulse oximetry (SpO2) was developed on the basis of a Nonin OEM III oximetry module and a radiofrequency transceiver. The electronic unit and the power supply was integrated into a specially designed infant shoe named BBA bootee. Clinical evaluation of the system revealed a good agreement between the pulse oximetry data transmitted by the BBA bootee and those recorded simultaneously by a reference monitor. The comparative data collected in 39 babies yielded a mean (bias SD) value of (−1.1 1.9)% for SpO2 and (−2 8) beats per minute for heart rate. Use of an integrated accelerometer/actimeter to reduce the motion artifact is addressed as well as ergonomics of the sensor-supporting garment.

Signal-to-noise ratio of a biopotential measurement system is determined not only by the electronics of the measurement device but also by the electrochemical noise of the electrode-electrolyte interface. The intrinsic electrochemical noise depends on the electrode/electrolyte interface used. Noise of three metals suitable for implantation (Au, Pt and Stainless Steel) are being examined and referenced to Ag/AgCl commonly used as surface biopotential electrode material. Measurements of the electrochemical noise have been conducted in saline, PBS and SBF. Two different sizes of electrodes of every material have been used. Frequency bandwidth of 0.5 – 500 Hz was used in the measurements.

The study presented here aims at developing a PVDF sensor that can be inserted within a seat-cover fabric. The final objective is to measure accurately the mass of a passenger sitting in a car. The measurement method consists in analyzing the response under compressive stress of a PVDF disc using the resonant frequency of the material. The maximal phase of the sensor at resonant frequency is taken up for each constraint applied. The influence of external parameters (vibrations, temperature variations) has been studied. The position of the sensor in the seat-cover was given starting from an ergonomic study. Tests in real case were carried out. A linear variation of the phase has been shown only taking into account the mass of the subjects’ sitting position.

Celeritas is an artistic/scientific collaboration between the Tyndall National Institute (Cork), the Interaction Design Centre in Limerick, Cindy Cummings (Dance Artist, Cork) and Todd Winkler (Composer and Digital Artist, Brown University, USA). Research Teams at the Tyndall Institute are developing wireless sensor network nodes, also known as motes, and associated miniaturized sensors. Motes can be applied in many different domains, ranging from medical and environmental monitoring to everyday applications in ubiquitous computing. This project aims to apply Tyndall’s sensor system to create a wireless dance costume for audio/visual performance using inertial sensor monitoring technology.

Dancers could be regarded as experts on human movement, producing accurate and expressive actions that provide a rich testing ground for human-computer interaction. The collaboration will push the boundaries of both artistic practice and wearable mote technology, as we will adapt and apply the Tyndall mote platform in a prototype body suit embedded with sensors. Software developed by the Interaction Design Centre and Todd Winkler will then convert the movement information detected by the sensors into computer generated sounds and processed video images. This mapping allows the dancer (Cummings) to fuse aspects of the physical body with the extended possibilities of the electronic body.

This paper presents the hardware platform that has been developed for the Celeritas project. The system is based around the Tyndall 25mm Wireless Inertial Measurement Unit (WIMU) node. The WIMU system is designed for integration into a body suit, which is to be worn by the dancer, whose movements are extracted from the wearable network of sensors and processed by a high-level software system that connects to the dancer wirelessly.

For about 20 years, Cochlear Implants have successfully restored hearing in postlingual deaf or helped to acquire auditory communication in prelingual deaf patients. However, only in recent years Cochlear Implant manufacturers have implemented stimulating and recording protocols for evoked compound action potentials in order to assess auditory nerve function in relation to the implanted device. Along with the principles of stimulation and recording of TECAPs, two cases are presented which highlight their diagnostic value as well as their limitations. While the pooled data obtained from patients are yet too variable to serve as a predictor for individual stimulation strategies, TECAPs are highly useful in intraindividual follow-up of patients in whom Cochlear Implant stimulation currents may vary even after years and who require re-adjustment according to objectively registered neural responses.

The BASUMA (Body Area System for Ubiquitous Multimedia Applications) body sensor network will consist of several wearable or handheld wireless medical sensors and a PDA like base station. The sensors, designed in a way to minimize disturbance of the patient’s everyday life, will record important diagnostic parameters which are analysed on the base station. This will allow a continuous monitoring of the patient’s state of health outside the hospital. Sensors were developed for the measurement of electrocardiograms (ECGs), air and blood content of the thorax (thoracic impedance), body temperature, breathing rate and cough control, blood pressure, pulse rate, oxygen saturation, lung functions, reactive oxygen species (ROS) (exhaled H

2

O

2

), and lactate in breath condensate.

This paper addresses the design considerations of a novel earpiece photoplethymograph (PPG) sensor and its

in-situ

evaluation results. The device is encapsulated with multiple LEDs and photodiodes based on a reflective PPG design. A compact and low power circuitry was developed for signal control and conditioning. PPG signals with an averaged a.c./d.c. ratio of 0.001–0.01 and 10% relative strength (compared to finger-based approach) were recorded from the superior and posterior auricular skins. PPG signal integrity and heart rate detection accuracy were evaluated and the results showed that with adequate optical shielding and motion cancellation, the device could reliably detect heart rate both during rest and moderate exercise. The proposed sensor design is low power, easy to wear compared to conventional earlobe PPG devices.

This paper presents the development of systems monitoring human body motions and postures for clinical purposes. The hardware for these applications exploits a newly released commercial Micro-Electro-Mechanical (MEM) 3-axis accelerometer and a MEM 3-axis rate gyroscope being developed by HSG-IMIT. First the paper gives an overview of wearable 3-axis accelerometer systems and the corresponding data storage and transmission developed by ETB. The main part of the paper describes the design of the 3-axis rate gyroscope to be implemented in an Inertial Measurement Unit (IMU) for the body motion monitoring. Such health applications require IMUs of very low size to be able to fix the sensor cluster to the human being. This prevents the use of state-of-the-art IMUs implemented by three perpendicular orientated rate gyroscopes for this purpose. Thus a novel 3-axis rate gyroscope realised in one plane, on a single die, is being developed. With this device a reduction of the package size of multiaxial MEM sensors is achievable at least by a factor of ten. A further advantage of this approach is the reduced cost due to the omission of a spatial configuration and due to the small dimensions. Finally the synthesis of the 3-axis accelerometer and the 3-axis rate gyroscope to an IMU for monitoring 3D body segment orientation is addressed.

For preventive continuous cardiovascular monitoring an In-Ear-implemented system will be designed. Using the photoplethysmographic curve a personal In-Ear-sensor measures the physiological parameters. The data are transmitted to a portable data analyser via wireless network, automatically reviewed and in the case of passing a critical value, an alert will be sent out. This In-Ear acquisition is facilitated by a new remission sensor designed by CiS Institut für Mikrosensorik GmbH. This In-Ear sensor is the core component ensuring light emission and reception by planar set-up and therefore, it is capable to use the inherent advantages of this measuring site for vital signs acquisition. Initial results have supported this approach.

Tilt sensing is important for human body motion detection and measurement. Two tilt sensors are introduced in this paper, based on MEMS (Micro-electromechanical Systems) variable capacitors, and utilizing the gravitational effect on a suspended proof mass to detect inclinations. A symmetric comb structure with high aspect ratio is adopted to obtain high capacitance. The first device can achieve a full range (−90° to +90°) tilt angle detection and relax the high-resolution requirement of the readout system by its linear output characteristics. Based on the same concept, a novel inherently digital sensor is proposed. The digital signal can be read out without complex processing, and so low power consumption can be achieved. A fabrication process, and simulation and processing results, are presented.

This work describes the evaluation of a wearable plastic optical fiber (POF) sensor for monitoring seated spinal posture, and the development of a workstation interface for the posture monitoring system. A garment-integrated POF sensor was developed and tested on nine healthy subjects. Data from the wearable sensor were compared to data taken simultaneously from a marker-based motion capture system. Peak analysis of the resulting data show a mean value error of 0.64 degrees and a mean time error of 0.53 seconds. These results show that the wearable sensor approximates the accuracy of expert visual analysis, and provides enough accuracy of measurement to reliably monitor seated spinal posture. The initial development of the system hardware and software inte rface are also described.

The use of wearable sensors for home monitoring provides an effective means of inferring a patient’s level of activity. However, wearable sensors have intrinsic ambiguities that prevent certain activities to be recognized accurately. The purpose of this paper is to introduce a robust framework for enhanced activity recognition by integrating an ear-worn activity recognition (e-AR) sensor with ambient blob-based vision sensors. Accelerometer information from the e-AR is fused with features extracted from the vision sensor by using a Gaussian Mixture Model Bayes classifier. The experimental results showed a significant improvement of the classification accuracy compared to the use of the e-AR sensor alone.

We propose a methodology to determine the occurrence of falls from among other common human movements. The source data is collected by wearable and mobile platforms based on three-axis accelerometers to measure subject kinematics. Our signal processing consists of preprocessing, pattern recognition and classification. One problem with data acquisition is the extensive variation in the morphology of acceleration signals of different patients and under various conditions. We explore several effective key features that can be used for classification of physical movements. Our objective is to enhance the accuracy of movement recognition. We employ classifiers based on neural networks and k-nearest neighbors. Our experimental results exhibit an average of 84% accuracy in movement tracking for four distinct activities over several test subjects.

Using body sensor networks (BSN) in critical clinical settings like emergency units in hospitals or in accidents requires that such a network can be deployed, configured, and started in a fast and easy way, while maintaining trust in the network. In this paper we present a novel approach called

BLIG

(Blinking Led Indicated Grouping) for easy deployment of BSNs on patients in critical situations, including mechanisms for uniquely identifying and grouping sensor nodes belonging to a patient in a secure and trusted way. This approach has been designed in close cooperation with users, and easy deployment and ease of use are top priorities. We present an initial implementation and evaluation of the presented technology.

Approximate data collection is an important mechanism for real-time and high sampling rate monitoring applications in body sensor networks, especially when there are multiple sensor sources. Unlike traditional approaches that utilize temporal or spatio-temporal correlations among the measurements of the multiple sensors observing a physical process to reduce the communication cost, in this paper we explore the idea of assigning different context-dependent priorities to the various sensors, and allocating communication resources according to data from a sensor according to its priorities. Specifically, a higher number of bits per sample is allocated to sensors that are of higher priority in the current context. We demonstrate that the proposed approach provides accurate inference results while effectively reducing the communication load.

We describe the activity recognition component of the Soldier Assist System (SAS), which was built to meet the goals of DARPA’s Advanced Soldier Sensor Information System and Technology (ASSIST) program. As a whole, SAS provides an integrated solution that includes on-body data capture, automatic recognition of soldier activity, and a multimedia interface that combines data search and exploration. The recognition component analyzes readings from six on-body accelerometers to identify activity. The activities are modeled by boosted 1D classifiers, which allows efficient selection of the most useful features within the learning algorithm. We present empirical results based on data collected at Georgia Tech and at the Army’s Aberdeen Proving Grounds during official testing by a DARPA appointed NIST evaluation team. Our approach achieves 78.7% for continuous event recognition and 70.3% frame level accuracy. The accuracy increases to 90.3% and 90.3% respectively when considering only the modeled activities. In addition to standard error metrics, we discuss error division diagrams (EDDs) for several Aberdeen data sequences to provide a rich visual representation of the performance of our system.

Dietary behaviour is an important lifestyle aspect and directly related to long-term health. We present an approach to detect eating and drinking intake cycles from body-worn sensors. Information derived from the sensors are considered as abstract activity events and a sequence modelling is applied utilising probabilistic context-free grammars. Different grammar models are discussed and applied to dietary intake evaluation data. The detection performance for different foods and food categories is reported. We show that the approach is a feasible strategy to segment dietary intake cycles and identify the food category.

It is widely known that (FHSS) frequency-hopped spread spectrum is the mostly used spread technique which uses multilevel modulation and cover both infrastructure field and Ad-Hoc networks. This paper presents the architecture and implementation of a FHSS RF radio transmitter. The complete radio transmitter chain was implemented in which a direct digital frequency synthesize (DDFS) was used for fast frequency hopping and FSK modulation. The radio transmits data in the 900 MHz ISM band at a rate of up to 160 kbps using frequency shift keying (FSK) modulation with frequency hopping at a rate of up to 80 khops/sec. The transmitter is also designed to take part of a detailed transceiver chain, to be modular and reusable in other wireless communication systems.

This survey paper serves as an introduction to the challenges and needs related to wireless personal telehealth systems and provides an overview of ongoing activities in industry and various communication standards, aiming to enable plug-and-play interoperability. Specifically, we address a recently founded industry consortium, the Continua Health Alliance, and ongoing standardization efforts within the family of ISO 11073/IEEE 1073 standards, the Bluetooth SIG, and the ZigBee Alliance.

This paper discusses the propagation channel between two half-wavelength dipoles placed near a human body. Different parts of the body are investigated separately. Statistical properties of the wireless on-body channel have been investigated. Path loss parameters and time domain channel characteristics are extracted from the measurement data. Path loss models for the arm and torso have been derived. A comparison with a path loss model near a flat, homogeneous medium has been made.

The Healthy Aims project,

www.healthyaims.org

, is a 23M€ four year project funded under the IST FP6 programme, with 16M€ combined funding from the EU and Switzerland. The project goal is to develop a number of intelligent medical implants and diagnostic systems integrating a range of underpinning Micro and Nano technologies. After three years a number of results are already available, disseminated to over 400,000 people worldwide and summarised in this paper.

The project has 25 partners from ten EU countries, including seven SMEs, six clinical partners, five LEs and seven academics and research groups. This combination of disciplines has enabled the Consortium to design, fabricate and carry out laboratory and clinical trials on the range of medical products under test. It has also enabled the project to set realistic specifications for core components that form an integral part of the implants, for example the implantable battery and the wireless data communications system. A further valuable benefit is that the micro and nano structures are designed specifically for use in an application, so again the specification requirements are clearly defined.

Functional reanimation of a paralyzed limb requires stimulation and sensing in multiple sites distributed throughout the limb. This paper describes a communication and control system for BION2™, a biomimetic system based on wireless, injectable neuromuscular implants that can detect the user’s intention, activate muscles and monitor the ensuing movements to support distributed feedback control with delays comparable to those achieved by the spinal cord. Various design features optimize the safety and efficacy of a system that must be reconfigurable to achieve many different clinical requirements despite highly constrained power and bandwidth.

In sensor networks, energy availability is often viewed as a constraint in the long-term sustainability of the network. Duplicate message suppression and message aggregation are high-level layer approaches to reduce power consumption. Body sensor networks have similar constraints however the use of aggregator nodes may increase the heat dissipation in the tissue surrounding these nodes. Due to additional processing, this additional heat may raise the tissue temperature to a level damaging to the individual. We propose a model of data reporting that balances power consumption with body temperature thresholds. We present a generic algorithm along with implementation analysis via simulation.