Pattern Recognition ofEEG Signals During MotorImagery

Electroencephalograph (EEG)recordings during theright andtheleft handmotorimagery canbeusedtomoveacursor toatarget onacomputer screen. SuchanEEG-based brain-computer interface (BCI) canprovide anewcommunication channel toreplace animpaired motorfunction. Itcan beusedbye.g., handicap users withamyotrophic lateral sclerosis (ALS). Theconventional method purposes therecognition oftheright handandtheleft handmotorimagery. Inthis study, statistical pattern recognition method based onARmodelisintroduced todiscriminate theEEGsignals recorded during theright hand, theleft handandtheright foot motorimagery. Andtherecognition ofthehand motorimagery andthefoot motorimagery also iscarried out. Finally, theeffectiveness ofourmethod isconfirmed through theexperimental studies.