The number of Americans unable to live independently in their homes due to cognitive or physical impairments is rising significantly due to the aging of the population. There is a currently a fundamental gap in the knowledge base concerning how to apply machine learning technologies to improve health monitoring and how to harness these technologies to implement interventions designed to sustain independent living. The long-term objective of this work is to improve human health and impact health care delivery by developing smart environments that aid with health monitoring and intervention. The objective of this particular application is to design, implement, and evaluate technologies for assessing everyday functional limitations and for providing automated intervention strategies for persons with early-stage dementia. To most people home is a sanctuary, yet today those who need special care, predominantly older adults, must leave home to meet clinical needs. The central hypothesis is that many older adults with cognitive impairment can lead independent lives in their own homes with the aid of automated assistance and health monitoring. The rational for the proposed work is that smart environment technologies can improve quality of life and health care for older adults who require assistance with everyday functional activities and reduce the emotional and financial burden for caregivers and society. Guided by strong preliminary data and a partnership between computer science and clinical neuropsychology researchers, our central hypothesis will be tested by pursuing the following specific aims: (1) Design software algorithms that use smart environment data to recognize complex everyday activities in real-world settings, (2) Use smart environments to automate functional health assessment and to examine the ecological validity of laboratory-based measures, (3) Design automated reminder and prompting-based interventions to aid with everyday activity completion, and (4) Analyze correlations between everyday behavioral patterns and physiological data. The proposed work is innovative because it defines methods of detecting and coping with aging, early dementia and disabilities in our most personal environments: our homes. The proposed work is significant because it provides the basis for automated, robust functional assessment of individuals with cognitive limitations and of intervention strategies designed to improve functional independence for these individuals. Rather than relying on self-reporting by the patient or by an informant who may or may not spend extended time with the patient, smart home technologies will allow us to identify functional deficits that impede a patient's ability to maintain independence in their home as they begin to occur, and to extend independent living at home by intervening in a real world setting.

Public Health Relevance

The proposed studies represent the first application of smart environment and machine learning technologies to the problem of in-home functional performance assessment. The result of this research will be software algorithms embedded in everyday environments that provide more reliable and accessible tools for measuring functional limitations, and that lay the foundation for creating and assessing tools that sustain or improve the ability to accomplish everyday tasks of living. Once such technologies become available, there is the likelihood that they can provide an ecologically valid method for monitoring an individual's everyday functional status and for extending the amount of time individuals can live independently in their own homes through the use of automated and reminder-based intervention.

Agency
National Institute of Health (NIH)
Institute
National Institute of Biomedical Imaging and Bioengineering (NIBIB)
Type
Research Project (R01)
Project #
5R01EB009675-04
Application #
8432827
Study Section
Biomedical Computing and Health Informatics Study Section (BCHI)
Program Officer
Korte, Brenda
Project Start
2010-03-08
Project End
2014-02-28
Budget Start
2013-03-01
Budget End
2014-02-28
Support Year
4
Fiscal Year
2013
Total Cost
$291,146
Indirect Cost
$87,203
Name
Washington State University
Department
Engineering (All Types)
Type
Schools of Engineering
DUNS #
041485301
City
Pullman
State
WA
Country
United States
Zip Code
99164
McAlister, Courtney; Schmitter-Edgecombe, Maureen (2016) Executive function subcomponents and their relations to everyday functioning in healthy older adults. J Clin Exp Neuropsychol 38:925-40
Roy, Nirmalya; Misra, Archan; Cook, Diane (2016) Ambient and smartphone sensor assisted ADL recognition in multi-inhabitant smart environments. J Ambient Intell Humaniz Comput 7:1-19
McAlister, Courtney; Schmitter-Edgecombe, Maureen (2016) Cross-sectional and longitudinal analyses of everyday memory lapses in older adults. Neuropsychol Dev Cogn B Aging Neuropsychol Cogn 23:591-608
McAlister, Courtney; Schmitter-Edgecombe, Maureen (2016) Everyday functioning and cognitive correlates in healthy older adults with subjective cognitive concerns. Clin Neuropsychol 30:1087-103
Dawadi, Prafulla Nath; Cook, Diane Joyce; Schmitter-Edgecombe, Maureen (2016) Automated Cognitive Health Assessment From Smart Home-Based Behavior Data. IEEE J Biomed Health Inform 20:1188-94
Thomas, Brian L; Crandall, Aaron S; Cook, Diane J (2016) A Genetic Algorithm Approach to Motion Sensor Placement in Smart Environments. J Reliab Intell Environ 2:3-16
Mcalister, Courtney; Schmitter-Edgecombe, Maureen; Lamb, Richard (2016) Examination of Variables That May Affect the Relationship Between Cognition and Functional Status in Individuals with Mild Cognitive Impairment: A Meta-Analysis. Arch Clin Neuropsychol 31:123-47
Dawadi, Prafulla N; Cook, Diane J; Schmitter-Edgecombe, Maureen (2016) Modeling Patterns of Activities using Activity Curves. Pervasive Mob Comput 28:51-68
Das, Barnan; Cook, Diane J; Krishnan, Narayanan C et al. (2016) One-Class Classification-Based Real-Time Activity Error Detection in Smart Homes. IEEE J Sel Top Signal Process 10:914-923
Feuz, Kyle D; Cook, Diane J (2015) Transfer Learning across Feature-Rich Heterogeneous Feature Spaces via Feature-Space Remapping (FSR). ACM Trans Intell Syst Technol 6:

Showing the most recent 10 out of 52 publications