Congratulations to Ratree Wayland, Kevin Tang, and their collaborators Karen Hegland and Yonghee Oh (College of Public Health and Health Professions), who were recently awarded a UF Research Opportunity Seed Fund for their project entitled, “Demystifying perceptual evaluations of parkinsonism sub-categorizations using machine learning”. The abstract for their project is below.
Parkinsonism refers to multiple neurodegenerative diseases that are characterized by common features of resting tremor, muscular rigidity, bradykinesia (slowed movement) or akinesia (no movement), and postural instability. Parkinson’s disease (PD) is the most common form of parkinsonism, followed by atypical parkinsonism such as progressive supranuclear palsy (PSP) and multiple systems atrophy (MSA). Accurate differentiation among these diseases is critical in developing a management plan, as PSP and MSA involve more wide-spread nervous system damage, respond poorly and with more side effects to traditional PD medications, and are much more quickly progressive than PD. Unfortunately, there are not currently any easily applicable diagnostic test(s) or biomarker(s) for PD, PSP, or MSA, and the diagnosis is on motor-based clinical criteria, which often leads to incorrect diagnosis of PSP and MSA as PD, especially early in the disease process. Recent studies have shown that quantitative assessment of various acoustic and rhythmic measures of speech can accurately differentiate speakers with different motor speech disorders. Given the different neuropathology of PD, PSP, and MSA, and the resulting distinctive characteristics of speech, we propose to develop a robust classification model to differentiate patterns and severity of dysarthric speech associated with PD, PSP, and MSA. This is an important first-step for our long-term goal of creating a robust set of speech-based biomarkers of these different diseases that may be able to detect atypical parkinsonism early in the disease.