Master’s Presentation for Hope Haygood

Friday December 9, 2022, 2:00 pm – 4:00 pm in Student Commons Building 4017

Title: Clock Drawing Test Project

Committee: Dr. Erin Austin (Chair, Advisor), Dr. Kendra Young, Dr. Greg Kinney, Dr. Yaning Liu

Abstract: According to the World Health Organization, “There are more than 55 million individuals worldwide that are living with Dementia” (World Health Organization). The Clock Drawing Test is a simple test that screens for Dementia in which a patient is given a pre-drawn circle and is asked to create a clock within the circle showing the time as ten minutes past eleven. The patient’s clock is scored on a binary basis (either being at High-Risk for Dementia (a score of 0) or not being at High-Risk for Dementia (a score of 2)) manually by psychologists. The Mini-Cog test is another tool utilized to screen individuals for Dementia that involves a scored Clock Drawing Test and a 3-item recall test for memory (Mini-Cog). According to the Center for Disease Control, "Chronic Obstructive Pulmonary Disorder (COPD) refers to a group of diseases that cause airflow blockage and breathing-related problems” (Center). When someone has COPD, they are at an extremely elevated risk for Dementia (Faris). People with COPD have lower oxygen levels while having higher carbon dioxide levels which causes harm to the brain (Faris). COPDGene aims to find out why an individual is more likely to develop COPD than someone else (COPDGene). The COPDGene Study has collected data from over 10,000 participants that include their scores on the Clock Drawing Test along with their scores on the Mini-cog test (COPDGene); this data was in the Clock Drawing Test Project. A previous researcher has created a machine learning algorithm that can manually score the Clock Drawing Tests (either being at High-Risk for Dementia or not at High-Risk for Dementia). We will complete an in-depth study on the High-Risk for Dementia Clock Drawing Test Images. Currently, an individual’s score on the Clock Drawing Test is only an early indicator for dementia; not specifying which type of dementia, or letting an individual know if they are at High-Risk for any other diseases linked to Dementia such as (COPD). There are 100 different types of Dementia (Adi - types of dementia). The Clock Drawing Test Project is a Machine Learning project designed to perform an in-depth study on the at High-Risk for Dementia Clock Drawing Test images by pixelating the images to be put into a Convolution Neural Network to then perform Clustering Algorithms (K-Means and K-Medoids) on the convoluted pixelated image data. Furthermore, we will investigate the clusters obtained from the Clustering Algorithms to determine if the patterns within the clocks obtained by the Clustering Algorithms could be an early indicator for COPD outcomes. The importance of the Clock Drawing Test Project is that it could potentially allow for the Clock Drawing Test to not only be an early indicator of Dementia but also an early indicator of COPD.