Classification of Sleep Staging For Narcolepsy Assistive Device
MetadataShow full item record
Narcolepsy is a chronic neurological disorder caused by the brain’s inability to regulate sleep wake cycles normally . Narcoleptic patients feel overwhelmingly tired and sleepy. They do not have the ability to carry out normal day time activities, such as work or study, hence proper treatment is essential. In order to provide an accurate diagnosis of the sleep disorder, physicians must analyze the sleep stages of the patient. Sleep staging analysis is the process of extracting sleep information with brain signals known as electrophysiological signals. There are three major electrophysiological signals: Electroencephalograms (EEG), Electro-oculograms (EOG), and Electromyograms (EMG). Through the three signals, physicians and technicians can classify the sleep stages. Although all three signals are important, most physicians and researchers agree that 95% of information can be extracted from EEG signal. With the current technology, patients must go to the hospital and sleep there over night to perform the sleep stage studies. Electrodes are placed on their scalp, eyelids and skin for this examination. Often patients feel that it is very inconvenient and time consuming. Moreover, the technicians are prone to make human errors during the classification of the sleep stages. These errors are a result of fatigue that the technicians experience while doing the long process of classification of the sleep stages, and the complexity and ambiguity of the rules to determine the sleep stages. Our research group has worked together to construct a portable device that will provide advice to the narcolepsy patient for activity planning and medication dosage. In addition, it provides fore-warning to the patients prior to an narcoleptic attack. This device will also perform real-time sleep analysis and alertness assessment through processing of electroencephalogram (EEG) signal. The classification accuracy is extremely important to the development of this device. With high accuracy of the classifier, treatment for the patients can be determined more accurately by the physicians. As a result, the main purpose of the research presented in this thesis is to analyze different classification methodology and to optimize the parameters of each technology to obtain the optimal sleep stage classification results. The thesis will also present the description of the portable device and its components used for the development of the prototype.