Work

運動想起脳波の特徴抽出に関する研究

Year: 2013
Type: dissertation
Abstract: This study proposes signal processing methods to presume or classify brain state of a human by using signals observed by electrodes installed on a scalp (EEG; Electroencephalogram).The issue of classifying the brain states from EEG signals is important for realization of brain computer/machine interface (BCI/BMI) and its applications (e.g.rehabilitation).Target states of the brain to be classified in the study are brain states when the human is imaging movements of he/she own hands and feet.The task performing the imaginations is called motor-imagery (MI) task.The goal in this classification is to classify EEG signals into classes corresponding to motor-imagery parts of body (e.g.right/left hand, feet) As features in the EEG signals associated with the MI tasks, the variation in energy of certain bands observed around motor cortex is known.For extracting the features, we can use a spatial filter that weights with different coefficients to each electrode, a frequency filter that extracts the certain bands, and a time window that removes periods not including feature components.In many cases, the parameters such as the coefficients of the filters and the window are empirically determined based on knowledge of neuroscience.For instance, the electrodes are installed around motor cortex, the passband of the frequency filter is set to 7 to 30 Hz including the band of the so-called mu rhythm the energy of which decrease by MI tasks, and the time window extracts signals observed from a start cue having i This work would not have been possible without the guidance of my academic supervisor, Dr. Tohihisa Tanaka.
Language: English
Open Access status: green
OpenAlex ID: W7163300320
Cites: 123
Cited by:
Related to: 10