A longitudinal study on identification of wandering patterns and their trends in elders with dementia residing in assisted living facilities

Abstract

Dementia is a chronic disease which affects the brain and its ability to function normally. It leads to the decline of brain functionalities which hampers its abilities in making judgments, processing language, planning, and sometimes leads to abnormal behavior. Wandering is a significant behavioral problem in Person with Dementia (PWD) and is a heavy burden on caregivers in residential and nursing homes. There have been various cases reported of elopement, getting lost and some serious accident arising out of abnormal wandering behavior. Wandering can be manifested in various forms. Martino-Saltzman in 1991 classified independent travel into four patterns, namely direct, random, pacing and lapping. Out of these four patterns, random, pacing and lapping are considered as wandering. The study of wandering in dementia takes this as reference topology for classification. The objective of this thesis is to create an end-to-end framework for wandering management. The author aims to develop set of algorithms which can be used for the indoor as well as the outdoor environment. The author validated the algorithm which resulted in an overall accuracy of 90%, which is more than any other previous methods used for such hybrid condition. The author also studied the qualitative as well as quantitative aspects of wandering. Qualitative analysis of travel patterns and its variability over a longer period is indicative of the cognitive status of PWD. These measures can be used to alert about the deteriorating condition of patients and can be used to ascertain the usefulness of certain treatment. This system can also be used in a rehabilitation center to monitor progress of the patient.