DETECTING DYSKINESIA AND TREMOR IN PEOPLE WITH PARKINSON'S DISEASE OR ESSENTIAL TREMOR DURING ACTIVITIES OF DAILY LIVING USING BODY WORN ACCELEROMETERS AND MACHINE LEARNING ALGORITHMS
Darnall, Nathaniel David
MetadataShow full item record
DETECTING DYSKINESIA AND TREMOR IN PEOPLE WITH PARKINSON'S DISEASE OR ESSENTIAL TREMOR DURING ACTIVITIES OF DAILY LIVING USING BODY WORN ACCELEROMETERS AND MACHINE LEARNING ALGORITHMSAbstractBy Nathaniel David Darnall, Ph.D.Washington State UniversityDecember 2014Chair: David C. LinParkinson's disease (PD) is a progressive neurodegenerative disorder that causes fluctuating motor deficits such as akinesia, bradykinesia, impaired balance, and tremor. Time periods characterized by severe deficits are referred to as "OFF" periods, while periods of relatively normal function are considered "ON" periods. Clinicians treat these deficits through the combined administration of Carbidopa/Levodopa medication, dopamine agonist medication, and deep brain stimulation. While motor deficits can be reduced, overmedication or overstimulation can cause dyskinesia, an involuntary, rhythmic or choric, exaggeration of movements. Clinicians assess the occurrence of motor deficits and dyskinesia, in part, by asking patients to retrospectively self-report how frequently these periods occurred over the several months prior to the clinical visit. This method is subject to recall bias. To augment the clinical assessment, several systems have been developed to provide clinical ratings from body-worn sensor data using computational algorithms. However, these systems place a substantial time burden and inconvenience to both clinicians and patients. Our ultimate goal is to develop an objective system that continuously identifies tremor, dyskinesia, and non-dyskinesia periods that occur during activities of daily living without placing a time burden on a clinician. We hypothesize that we can classify body-worn accelerometer data into tremor, dyskinesia, and non-dyskinesia periods using signal analysis, feature extraction, and machine learning algorithms (MLAs). This research will focus on three specific aims:1.Classify kinematic data collected during clinical assessment tasks onto tremor severity ratings using machine learning algorithms.2.Develop a system that classifies features derived from body-worn accelerometer data as dyskinesia presence that was determined from visual observation of participants performing unconstrained activities of daily living.3.Determine factors that would generalize the dyskinesia detection system for continuous in-home use.