Georgia Institute of Technology
Mild Cognitive Impairment (MCI) is an early-stage decline in memory, language perception, visual perception, and/or other cognitive abilities that is greater than expected for an individual's age and education level, but that does not greatly impact their ability to perform everyday activities. Research has shown that this condition is present in 3-19% of people over the age of 65. As a part of the Cognitive Empowerment Program (CEP), we have participants with MCI enrolled in a study allowing us to collect daily data by installing sensors within their homes. To detect progression early and take timely preventive measures, our project aims to understand if the progression of illness or decline to dementia can be detected via this collected data. One challenge faced when collecting this data is due to the unreliable nature of this data and pre-processing it correctly is a huge component the team is focussing on. In addition, the team has engineered differen t features to define indicators of changes in activity within the household and selected a baseline period of 2 months for these features aiming to identify the household normal. For a selected set of households (with relatively reliable data and no external factors like pets), deviation from this defined normal was checked over time and two homes had significant changes from their baselines. Clinicians validated that these homes indeed had members who showed some indications of decline due to various factors like observed hospitalizations. Currently, this analysis is being expanded to validate it in unseen scenarios and how it reacts to the addition of external factors. The eventual goal of this research is to create a set of features that can be utilized to build a robust model to predict future changes reflective of potential decline within MCI members.