A Mind Controlled Mouse for Elderly People with Multiple Disabilities using Discrete Wavelet Transforms and Independent Component Analysis

摘要

According to a statistical survey made by World HealthOrganization (WHO) India suffers from highest number of roadaccidents and out of which more than ten percent of them are proneto head injuries. This scenario leads to patient’s death or make thevictim to become comatose. Also, many different disorders candisrupt the neuromuscular channels through which the braincommunicates with and controls its external environment. Brainstemstroke or spinal cord injury, cerebral palsy, muscular dystrophies,multiple sclerosis and numerous other diseases impair the neuralpathways leading to communication and control which make thevictims intellectually or physically disabled. Most often, thecommunication for paralyzed people is established by using a BrainControl Interface (BCI). Most of the existing systems hadexperimented Brain Computer Interface either with animals orhealthy human beings. But, this paper focuses on movements of themouse cursor controlled by a person with multiple disabilities. Themouse cursor movement would further be used by the disabled personto have a communication with his caretaker by means of the softwaredeveloped by us. The proposed system uses discrete wavelettransforms for de-noising the muscular and cardiac signals. Anindependent component analysis is performed in order to extract thebeta rhythms from the EEG signal. The mouse control is achieved byinterfacing the mouse with a microcontroller which receives theoperating voltages from the Data Acquisition System (DAS) whichacquires and conditions the EEG signals coming from the user brain.The proposed system is tested on several young and elderly personsand is found to be working with more than 95% accuracy.