Understanding Activity Patterns in Dementia Patients Using Wearables

Poster Number: P128
Presentation Time: 05:00 PM - 06:30 PM

Abstract Keywords: Mobile Health, Tracking and Self-management Systems, Ubiquitous Computing and Sensors, Aging in Place
Primary Track: Applications
Programmatic Theme: Consumer Health Informatics

Wearables enable activity tracking for dementia patients, with many employing step-count averages over multiple days to analyze trends. This study examined the impact of different time windows for averaging step counts on preserving activity variability based on the step counts of 7 days collected using Fitbit Inspire 2 with a sliding window approach. Our analysis suggests that shorter windows captured natural day-to-day variations, whereas longer windows over-smoothed fluctuations, obscuring significant changes.

Speaker(s):
Saitejaswi Cherukupalli, Master of Science
Indiana University

Author(s):
Saitejaswi Cherukupalli, Master of Science - Indiana University; Charanjit Kaur, Master's in Health Informatics - Indiana University Purdue University; Jean-Francois Daneault, PhD - Rutgers University; Hee Tae Jung;