Quantification of Interlocking Pentagon Copying

• Activity Number: 71 - Statistical Methods for Personalized Medicine
• Type: Contributed
• Date/Time: Sunday, July 29, 2018 : 4:00 PM to 5:50 PM
• Sponsor: Biometrics Section
• Author(s): Namhee Kim* and Timothy Truty and Lisa L Barnes and David A Bennett and Melissa Lamar
• Companies: Rush University Medical Center and Rush University Medical Center and Rush University Medical Center and Rush University Medical Center and Rush University Medical Center
• Keywords: Hierachical clustering; histogram oriented gradients; interlocking pentagons; Alzheimer's disease

Abstract

Interlocking pentagons copying (IPC) is a part of the Mini-Mental State Examination (MMSE), which is widely used for dementia screening. In IPC, participants are asked to copy a given picture of two interlocking pentagons. While binary (0=incorrect, 1=correct) scoring of the MMSE IPC item is the standard approach, this gross assessment may overlook subtle alterations in performance that may signal risk for cognitive impairment. We propose a series of algorithms to extract quantitative measures from IPC data, e.g. total and individual areas of the two pentagons, intersecting area. Algorithms adopted for this study include histogram oriented gradients for corner and line detection from individual drawings of IPC on paper, hierarchical clustering for separation of digitally regenerated two interlocking pentagons, and Monte Carlo integration for calculation of areas and other desired measures. We also propose statistical procedures to measure goodness of performance of our approach. Finally, we demonstrate the procedure with over 10,000 IPC's collected in longitudinal studies at the Rush Alzheimer's Disease Center.