Detection of system change for the brain activity by EEG monitoring
| Project/Area Number |
16560353
|
|---|
| Research Category |
Grant-in-Aid for Scientific Research (C)
|
| Allocation Type | Single-year Grants |
|---|
| Section | 一般 |
|---|
| Research Field |
System engineering
|
|---|
| Research Institution | The University of Tokushima |
|---|
Principal Investigator |
NAGASHINO Hirofumi The University of Tokushima, Faculty of Medicine, Professor, 医学部, 教授 (40035655)
|
|---|
| Co-Investigator(Kenkyū-buntansha) |
KINOUCHI Yohsuke The University of Tokushima, Institute of Technology and Science, Professor, ソシオテクノサイエンス研究部, 教授 (80035807)
AKUTAGAWA Masatake The University of Tokushima, Institute of Technology and Science, Lecturer, ソシオテクノサイエンス研究部, 講師 (90294727)
|
|---|
| Project Period (FY) |
2004 – 2006
|
| Project Status |
Completed (Fiscal Year 2006)
|
| Budget Amount *help |
¥3,800,000 (Direct Cost: ¥3,800,000)
Fiscal Year 2006: ¥700,000 (Direct Cost: ¥700,000)
Fiscal Year 2005: ¥700,000 (Direct Cost: ¥700,000)
Fiscal Year 2004: ¥2,400,000 (Direct Cost: ¥2,400,000)
|
|---|
| Keywords | EEG / brain / brain activity / system change / time series signal / neural networks / backpropagation learning / coupling weight / 誤差逆伝搬学習 / 時系列 |
|---|
| Research Abstract |
1. Analysis of system change detection using artificial neural networks
We elucidated that moving average type artificial neural networks have the ability to predict time series generated by linear or nonlinear systems and track the states of the systems that are nonstationary. We train the networks with the time series signal and examine the coupling weight space to extract information about the signal generator. We defined a novel feature to measure the separation of two arbitrary states in the coupling weight space and use it to detect the state changes of the generating system.
2. Development of the detection system of brain activity system change by EEG measurement
We applied the detection system to brain activity change by clinical EEG measurement. Clinical data obtained from patients undergoing carotid endarterectomy of the brain showed that EEG could be modeled by the proposed method. Moreover, we developed the methods of removing the artifacts by blink etc. using independent component analysis or artificial neural networks.
3. Development of an EEG monitoring system
We developed the EEG monitoring system by EEG measurement from the electrode placement based on the international 10-20 system. It is an integrated EEG analysis software. It consists of data acquisition, waveform memory management and data analysis components. We realized the real time processing of the above system change detection using neural networks as well as DRT (deviation ratio topography), frequency spectrum and topography changes of each frequency components.
|
Report
(4 results)
Research Products
(31 results)
