Back to search

Aristotle University Of Thessaloniki (AUTH)

Ανάλυση εγκεφαλογραφήματος με προηγμένες μεθόδους επεξεργασίας σήματος για την υλοποίηση διεπαφής εγκεφάλου - υπολογιστή

Abstract

dc:description

Brain computer interface systems are designed to recognize the will of the user and convert it into a control signal by analyzing information recorded directly from the brain. These systems are used to improve the quality of life for people with disabilities, mainly those suffering from the locked in syndrome. For those people such systems are the only available communication channel. They can also be used by people suffering from quadriplegia for communication and to a lesser extent to control prosthetic parts. To implement such systems, EEG has been widely used due to its non invasive nature, its ease of recording, and the comparatively low price of the necessary equipment. The practicality of existing systems is limited both by the low recognition rate and the relatively high rates of incorrect recognition. These two factors prohibit the widespread acceptance of the system by the majority of the users. This thesis deals with the creation of a system of brain-computer interface by using advanced signal processing methodologies to process the EEG in order to recognize faster and more reliably the will of the user. For this purpose, the Hilbert-Huang Transform was used which was originally designed for the analysis of nonstationary and nonlinear signals. The application of the Hilbert-Huang Transform combined with Independent Component Analysis algorithm created an composite algorithm that can detect the activation of consciously controlled brain centers. This algorithm works more reliably than existing methods on brain computer interfaces. The performance of the algorithm was verified by using databases created specifically for the benchmarking of brain-computer interfaces in a series of global competitions under the name "BCI Competition". Acknowledging the importance of the learning stage in the acceptance and integration into the daily life of a disabled person of an brain - computer interface, a very high precision algorithm which aims to minimize the learning curve was created in this thesis. This algorithm is based on the calculation of the amount of information transferred between different brain regions as determined by the locations where the EEG is recorded. In conclusion, the methodology developed in the context of this thesis is a combination of signal processing methods, such as the Hilbert - Huang transformation, the ICA algorithm and multidimensional directed information data. The accuracy of the proposed algorithms in estimating the will of the user is better if compared to the ones used today.

Degree

thesis:*
Grantor dc:publisher
Aristotle University Of Thessaloniki (AUTH)
Year dc:date
2010

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Panoulas, Konstantinos
  • Πανούλας, Κωνσταντίνος

Subjects

dc:subject × 12

Rights

Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/29822

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Panoulas, Konstantinos; Πανούλας, Κωνσταντίνος. Ανάλυση εγκεφαλογραφήματος με προηγμένες μεθόδους επεξεργασίας σήματος για την υλοποίηση διεπαφής εγκεφάλου - υπολογιστή. Aristotle University Of Thessaloniki (AUTH), 2010. http://hdl.handle.net/10442/hedi/29822