SJf_Wettbewerbs_Broschüre_2007 - Die Goldene Sonne am Calanda

Spektralanalyse des Elektroenzephalogramms zur Klassifizierung
durch ein Neuronales Netz
Introduction
Brain waves are electrical activities of the brain which are manifested as alternating potential
differences at the scalp surface and are also known as “electroencephalogram” (EEG). The
alternating potential differences encompass a range of typically ±75 microvolt under the
condition of quiet wakefulness, while body movements, in particular movements of legs and
arms, produce somewhat higher voltages.
Principal goals
In our project we aimed at distinguishing between several human body movements by analysing
and classifying the underlying brain electrical activities.
Methods
This task was tackled with the help of spectral analyses along with neural networks. Neural
network analysis models the human brain through neurons that are organized in layers and
interconnected to each other in a variety of ways. Thus, neural networks are able to process any
kind of input stimuli and to generate clearly identifiable responses to these stimuli.
Data material
Using an experimental design with seven different types of body movements (“experimental
conditions”) along with repeated assessments on the same individual at weekly intervals, we
were able to quantify brain waves, their within-subject fluctuations and their between-subject
variations. Having extracted the spectral information inherent in EEG time series, we applied
neural network analysis in order to construct classifiers that predicted the seven experimental
conditions from the EEG spectral values.
Results
After a suitable phase of learning, the algorithm yielded a final neural network that classified the
seven experimental conditions at a rate of 90% correctly classified probes.
Discussion
As our project was carried out in the sense of a pilot investigation, repeated assessments on
only two test persons are currently available, so that these results are preliminary and lack
general validity. On the other hand, our results can well be considered as a proof of principle,
thus stimulating further investigations that involve a much larger and more representative sample
of test persons. The program package developed within the scope of this project will greatly
facilitate attempts in this direction.
Physik / Technik
Raphael Blaser
8600 Dübendorf
1986
Kantonsschule Glattal Dübendorf
Würdigung
Diese interessante und vielseitige Arbeit
erlaubt die Erkennung verschiedener Bedingungen
(zum Beispiel Augenblinzeln)
anhand von Hirnstrommessungen. Dazu
haben Raphael Blaser und Eric Stassen
nicht nur ein Computerprogramm geschrieben,
sondern auch zahlreiche Experimente
durchgeführt. Die Arbeit zeichnet
sich durch eine systematische und
wissenschaftliche Herangehensweise
aus. Der Bericht ist anschaulich und gut
verständlich.
Prädikat
Hervorragend
Sonderpreis
„European Union Contest for young
Scientists“ in Valencia
Sonderanerkennung
Metrohm Stiftung Herisau
Experte
Dr. Volker Koch
AWK Group AG Zürich, Consultant
45
Eric Stassen
8600 Dübendorf
1987