BioMOBIUS Prototype Applications
BioMOBIUS software has been utilized by the TRIL Centre for a number of prototype applications both in clinical and home environments. A typical BioMOBIUS-type application comprising the software and SHIMMER wireless sensors described in this article is shown in Figure 9.
We now look at three TRIL Centre projects that have utilized the BioMOBIUS research platform.
Gait Analysis Platform
The BioMOBIUS platform has been used to develop a prototype gait analysis platform for use in a clinical environment and designed for early detection of gait and postural instability. The system is designed to unobtrusively capture gait parameters and physiological data. The platform comprises a number of integrated subsystems that provide a true six degrees of freedom analysis, linked to footfall data and visual gait data. Body-worn kinematic and sensor data are provided by SHIMMER sensors and via a Tactex pressure-sensitive floormat. Video data of the subject are provided via two Webcams. The hardware components are integrated into a BioMOBIUS high-level application that provides real-time visual feedback to clinicians. The system's user interface allows clinicians to select and adjust which data are collected and how the data are processed. The software encapsulates data acquisition and signal processing modules, and it allows customization of the sensors [68]. The system is being used in the TRIL Clinic in St James' Hospital, Dublin, Ireland to evaluate scientifically the impact in non-normal gait changes and associated falls' risks.
The MuSensor Project
The Tril Centre's MuSensor project aimed to develop a cheap, unobtrusive, home-based wireless sensor network to monitor an elder's gait velocity and doorway dwelling times over an eight-week period. The network was developed using BioMOBIUS software and SHIMMER's low-power 802.15.4 communications module and Passive Infrared (PIR) motion sensor daughterboards. Each activation of a PIR sensor was time-stamped and temporarily stored in the SHIMMER's RAM, and batches of PIR events were periodically uploaded via the 802.15.4 radio to a central data aggregator located in the home. Data were then transferred to a server via a 3G GSM network. The PIR sensor activation events were used to measure in-home velocity of fallers and non fallers to determine if non performing (i.e., measured outside a clinical environment) gait velocity of fallers and non fallers differs in the home.
Engineering Alertness
This project is focused on the development of, and validation of, a prototype BioMOBIUS application that can detect and correct lapses in alertness. The study tests the viability of a self-alert training (SAT) protocol supported by a BioMOBIUS software application with SHIMMER electrodermal activity (EDA) sensors for measuring exosomatic EDA.
A small DC current was passed across two electrodes placed on the skin, and the change in the resistance of the skin to the current was recorded as a function of increased sweat gland activity (Figure 10). For DC measurements, the current is kept constant and skin resistance is measured as changes in voltage.
The protocol implemented in the application is designed to assist older people in increasing attention and alertness levels in their own homes.
Developing the Ecosystem
BioMOBIUS software components can be downloaded for free from the BioMOBIUS website under license from the TRIL Centre. The website features a comprehensive set of documentation to support both application and block developers. A variety of tutorials and sample applications to assist new users are available. A user forum is also available where users can log questions for TRIL engineers. A number of international workshops have also been delivered to help with knowledge transfer in the research community.
The SHIMMER wireless sensor platform was licensed by Intel to Realtime Technologies in 2007. SHIMMER development kits are now commercially available from SHIMMER Research [42].
Conclusions and Future Work
BioMOBIUS software remains under active development. Version 2.0 of the platform was released in late summer 2009. It features improved stability, support for the X10 family of sensors, improved graphing capabilities in the GUI Developer, support for SQLite and MySQL databases, improvements to SHIMMER blocks, etc.
A number of new features are currently in development, including ODBC support, Continua-compliant device support and Web services data transports. New features and improvements will be available to download from the BioMOBIUS website.
The BioMOBIUS research platform represents a set of highly integrated software and hardware components developed by the TRIL Centre for prototype biomedical research applications. The software components are freely available for download to researchers. Hardware components, such as the SHIMMER wireless sensors platform can be purchased from third-party vendors. The platform is open and extensible, allowing users to develop new blocks for the graphical development environment or to integrate new hardware components into the platform. The TRIL Centre has demonstrated the use of the BioMOBIUS platform in a number of research projects including the development of prototype gait analysis platform and EDA monitoring applications. We believe that the BioMOBIUS research platform will prove to be a useful and versatile tool in enabling the biomedical research community. The prototype applications we developed with the BioMOBIUS research platform and described in this article are for research purposes only and are not commercially available.
Acknowledgements
We acknowledge the contributions of Barry Greene, Adrian Burns, Cliodhna Ní Scanaill, and the Digital Health Group Product Research and Incubation in the preparation of this article. We also recognize the TRIL Centre's researchers and their research programs, some of which are reported in this article.
