Wireless User Interfaces - site visit 2

Site visit in Tampere on 27 November 2003

The project studies lightweight wireless sensor technology that can be used for monitoring the human heart, brain activity, eye movements, and facial muscle activity. Such wireless technology could be utilized widely, for example, in developing hands free user interfaces for functionally challenged persons and in health care monitoring.

The first prototype of a wireless sensor measuring heart beats or electrocardiograms has been built out of existing components. The sensor works on battery (24h) and consists of a transmitter and an amplifier. Sampling is made at 500 Hz and the transmission range is about 20 m. Antti Vehkaoja is writing his M.Sc. Thesis about this first prototype which has been built out of existing components. Future plans include integrating the transmitter and the amplifier as well as the antenna on the same circuit. At the same time, inductive links and RF links will be studied to replace the battery and provide wireless power transfer to the sensor. New prototypes of the sensor will be built during 2004.

How small can a wireless sensor system be that measures bioelectric signals? And can a computer model predict the strength of the real bioelectric signal? Merja Puurtinen tries to answer these questions in her recent M.Sc. thesis "On the effects of the interelectrode distance and electrode size on bioelectric signal strength" (Tampere University of Technology 2003). She studied the effect of interelectrode distance on the ECG signals. By using a 2D model of the human chest she was able to deduce that reducing the distance between electrodes decreases the signal strength but also that the magnitude of the decrease depends on the electrode location. The results obtained from the model correspond fairly well with clinical results recorded from 236 patients. Further studies will now be undertaken with 3D models.

How do we recognise facial muscle activity from electromyograms? A simple and efficient way to do so is to use a multilayer perceptron (MLP) network. Results have been promising in a preliminary study where the data of several test persons were used. Electrodes are attached to several locations in the test person's face, e.g., to measure voluntary movements on the person's eyebrows (i.e. corrugator supercilii), cheek (i.e. zygomaticus major) or blinking. In a case where an MLP network was used to recognize eyebrow movements of a 6 persons, and decide if there was activity or not, over 96% of the data was correctly categorised. Further studies will be deducted on the eyebrow and cheek muscle constructing a similar network for recognising activity there and testing with more people (N=30). Pekka Niemenlehto is writing his M.Sc. Thesis on the subject and the thesis is expected to be ready in the spring of 2004.

An interactive physiological system would observe the person's physiological state and its changes and react to and give feedback regarding these changes. An interactive psychophysiological system contains a model combining physical and psychological states. This model is used as a basis for how the system reacts. Toni Vanhala uses an agent framework to implement such a psychophysiological system. The benefits of an agent framework is that it will be easier to adapt the system include new components (new sensors) and combine components. His M.Sc. Thesis will also be published later this year.

More information

A description of the research on the effects of interelectrode distance has recently been published in the International Journal of Bioelectromagnetism, see
Puurtinen, M., Hyttinen, J. and Malmivuo, J. Effect of interelectrode distance on ECG potentials -modelling approach vs. clinical data. International Journal of Bioelectromagnetism, 5(1): 147-149, 2003.

For further information, please contact the coordinator of the joint project, Assistant Professor Veikko Surakka at the University of Tampere, or the other researchers involved, as well as on the web page of the project.

 Project partners:

• University of Tampere (UTA), Department of Computer and Information Science
• Emotions, Sociality and Computing Group at TAUCHI: Assistant Professor Veikko Surakka, Researcher Toni Vanhala
• Signal Processing: Professor Martti Juhola, Researcher Pekka Niemenlehto
• Tampere University of Technology (TUT)
• Analysis and Modelling of Physiological Systems Group, Ragnar Granit Institute: Professor Jari Hyttinen, Researcher Merja Puurtinen
• Sensor Technology Group, Institute of Measurement and Information Technology: Professor Jukka Lekkala, Researcher Antti Vehkaoja

Project home page on this site

Greger Lindén
Programme Coordinator
Greger.Linden@cs.helsinki.fi