Laboratory for EEG Analysis

Laboratory for EEG Analysis at Center for Sensory-Motor Interaction, Department of Health Science and Technology, Aalborg University, focuses on quantitative methods for analysis of the electrical signals from the brain - the EEG.

Background

The brain controls our movements, level of consciousness, memory, senses etc.  The electrical activity from the brain – the ElectroEncephaloGram – (EEG), contain information of these processes and thereby it is possible to discriminate between disorders and normality. The information is hidden in the complex EEG generated by billions of neurones, but can be extracted by use of proper recording methodology and advanced signal-processing.
Laboratory for EEG Analysis focuses on quantitative methods for recording and analysis of the EEG for use in development of diagnostic and therapeutic methods and equipment. Also experimental research utilising the EEG to gain more knowledge of the function of our brain and body is done.

Main activities

Emphasis has been put on the development of new methods for diagnosis in Sleep,  Headache and Rehabillitation using advanced methods from signal-processing and informatics.
Methods and systems for diagnosis of sleep disorders in term of macro- and micro sleep descriptions have been developed and the methods are now used in clinically and experimentally oriented research. Especially the relation between sleep and pain has been investigated in collaboration with clinical partners from e.g. Aalborg Hospital.

Together with national and international partners methods for diagnosis and training in neurology has also been point of focus, a CD-based tutorial and an expert system for headache diagnosis is one of the results.

EEG and EP used for analysis in motor-control and in rehabilitation, is also central for the laboratory.
Especially focus is on Brain Computer Interfaces used for communication and control by movement impaired patients. Methods and prototypes using Steady State Visual Evoked Potentials (SS_VEP) has been develloped.
The laboratory also perform basic EEG-research with the aim to devellop more refined BCI-systems.

Future perspectives

The aim for the laboratory is to refine the analysis and use of the human EEG so it can be used not only for further diagnosis in neurological disorders and to gain more understanding of the human brain. But also be used further for direct control of computers and neural prosthesis and thereby be used as the optimal command-signal in rehabilitation.


Sleep research
 

 
 
 
 
 
 

Sleeplaboratory at Aalborg University in action 

Methods for automatic sleep analysis and diagnosis have been point of focus for some years now.
One of the outcomes of this was the devellopment of the sleep analyser: The Nightingale Polygraphic Sleep Analyser .

The department is equipped with a sleep lab with facillities for digital polygraphic recordings and analysis. This laboratorium is solely used for research.

In collaboration with a.o. Aalborg Hospital a large number of clinical oriented reseach projects using digital home-recordings has taken place. Methods for analysis of the sleep microstructure have been develloped and
used in analysis of the sleep microstructure in chronic pain disorders.

At present experimental research is done in the effort to establish knowledge of the interaction between sleep and pain and the effect of sleep deprivation on pain tolerence etc.

The department is an associate partner in the EU-Biomed II project called SIESTA  which is an "acronym" for: A New Standard for Integrating Polygraphic Sleep Recordings into a Comprehensive Model of Human Sleep and its Validation in Sleep Disorders. The SIESTA project runs from September 1997 to September 2000.

Publications

Contact:
Kim Dremstrup Nielsen, PhD
Asbjørn M. Drewes, Dr.Med

Danish Sleep Research Society


 

European Sleep Research Society
 
 
 
 
 
 


Decision support in Neurology and Sleep
 

AAU  particpates as an associated partner in the European Neurological Network (ENN). The project is running under EU's 4´th framework program in telematics. ENN is working within the three neurological domains: Sleep, epilepsy and headache. One of the tasks in ENN is development of a neurological database with acces via the internet. One of the purpose of this is development and validation of new decision suport systems to improve the capacity and quality of the diagnosis procedures. Another task in ENN is development of multimedia tutorial systems for teaching of GP's and and hospital physicians within the three mentioned domains.

The Diagnostic Headache Diary has been develloped in cooperation with Dr. Michael Russell, Rigshospitalet, Copenhagen. This decision support system is a part of the ENN Headcahe Tutorial System., a prototype of a multimedia tool for training of Neurologists.

Contact:
Kim Dremstrup Nielsen, PhD

Publications


Brain Computer Interface  / Motor Control

The Laboratory for EEG Analysis at Aalborg University focuses its research on quantitative methods of EEG recording and analysis for development of diagnostic and therapeutic methods and equipment. Experimental research is also conducted on respect to brain function associated to motor control with the purpose of gaining more knowledge in the perspective of rehabilitation.

Our focus on brain-computer interfaces (BCI) is so far directed towards the Steady-State Visual Evoked Potentials (SS-VEP) using a standard computer screen for stimulation. Accordingly, subjects are presented to a matrix of 3 by 3 squares, each flickering at a different frequency; thus, depending on which square the subject directs its attention it is possible to distinguish specific frequency components in the EEG signals recorded over the visual cortical area (occipital region). Feature extraction and classification are based on power spectral analysis and amplitude criteria, respectively.

This BCI based on SS-VEP appears to be a promising approach since accuracy of up to 92.8 % can be achieved with symbol signaling rate of 12 chars per minute. Besides, the required training time seems to be negligible and high detection speeds are possible, even though some control of eye movements is still needed. The current efforts on this kind of BCI are directed towards the implementation of the “multi-tap” alphabet (like those used in text messaging of mobile phones) in connection with predictive text input in order to speed up the information transfer rate and promote an efficient communication tool for patients.

In the lane of basic research we have been investigating the cortical modulation of movement-related parameters. The aim is to enhance the control of function electrical stimulation (FES) by integrating a BCI system capable of recognizing cortical activity related to e.g. amount of force, velocity and direction desired by the patient for a determined action. The control (and command) of these parameters may be of most importance when performing complex tasks like overcoming obstacles during walking, controlling pedals (e.g. driving a car), etc. Controllers of FES systems, provided with these movement-related cortical potentials (MRPs) through BCIs, would require simpler algorithms and strategies because part of the inputs would be already pre-programmed by the patient’s own preserved areas of the nervous system.

Considering the basic physical demands involved in essential motor functions associated to lower limbs, we have been studying the relationships associated to ankle plantar flexors. Accordingly, it has been possible to identify, for instance, that MRPs are function of rate of force development and force amplitude. Changes in directional orientation during gait initiation also seem to be directly modulated by MRPs mainly due to adaptations of the body geometry and muscle synergies. Moreover, the resemblance of this cortical modulation in imaginary movements has been also investigated, looking to the application of BCIs in patients which are deprived of normal neural pathways.  Last but not least, motor cortical modulation is being investigated on a single trial basis with techniques of feature extraction and classification involving signal-based wavelets.

A special attention has been given to spatial distributions of both MRP intensities and statistical significant differences of MRPs between different tasks. Hence, the identification of specific scalp regions associated to variations of movement parameters may help in the optimization of BCIs by allowing reductions in the number of recording electrodes without loss of information. In that sense, customized EEG caps with high density of electrodes over the motor cortical area have been employed in those experiments related to the investigation of motor cortical modulation.


 

Contact:
Kim Dremstrup Nielsen, PhD
Omar Nascimento, PhD
Alvaro Vuentes Cabrera, MsSc
Michael Voigt, PhD
Dario Farina, PhD
Saber S. Sami, MSc

 Publications .


Contact: kdn@hst.aau.dk


Local links:

[KDN's homepage] [SMI Home]
[Department Home] [Campus Home]