Detecting Multimodal Behaviors for Neurodegenerative Disease

Abstract

Neurodegenerative diseases such as Parkinson’s and Alzheimer’s are incurable and affect millions of people worldwide. Early diagnosis is critical for improving quality of life for patients. Current methods rely on the use of tests administered and evaluated by clinicians. The digital Symbol Digit Test (dSDT) is a novel cognitive test that aims to distinguish between individuals with normal and impaired cognitive abilities. This thesis will develop a framework for processing collected participant eye-tracking and handwriting data and show its use in detecting specific multimodal learning behaviors. Furthermore, this thesis will explore recommendations for working with eye-tracking systems and outline future steps towards developing a multimodal classification model to automate early diagnosis of neurodegenerative disease.