Analysis of Tonal Qualities within the Voice for Parkinson’s Solutions

by Amanda Norton

Submitted : Fall 2015

This project will examine two things. The first of which is compiled within a data set which measures voice changes versus the clinician based UPDRS score, (a score which measures cognition and progression of the Parkinson’s disease). Telemonitoring Parkinson’s disease is a non-invasive method for checking in on the patient and observing mental and physical progressions of the disease. A multiple linear regression will be performed to see which unique variables are significant, and subsequently a linear model will be created. The linear model can be used in the future to estimate the UPDRS score response based on independent variables found within the voice.

The second portion of the project will examine how exactly statistics, calculus, and Parkinson’s Disease are correlated. The multiple linear regression in itself can indeed be quite complex, but often it’s applications to actual mathematics can be overlooked. We will examine them, and thus apply them to the data set used which is regarding Parkinson’s disease.

[ Back ]