Related Experiment Video
Updated: Jun 13, 2025
07:31
A Computerized Functional Skills Assessment and Training Program Targeting Technology Based Everyday Functional Skills
Published on: February 13, 2020
6.9K
Predicting Adherence to Computer-Based Cognitive Training Programs Among Older Adults: Study of Domain Adaptation and
Ankita Singh1, Shayok Chakraborty1, Zhe He2,3
1Department of Computer Science, Florida State University, Tallahassee, FL, United States.
JMIR Aging
|September 16, 2024
Summary
This study used deep learning and domain adaptation to predict when older adults might stop cognitive training. The methods accurately forecast adherence lapses, paving the way for personalized support systems to improve engagement and cognitive health.
Area of Science:
- Gerontology and Cognitive Science
- Artificial Intelligence and Machine Learning
- Health Informatics
Background:
- Cognitive impairment and dementia present significant challenges for the aging population, impacting autonomy and quality of life.
- The growing elderly population strains healthcare and economic systems.
- Adherence to computerized cognitive training programs is often difficult for older adults.
Purpose of the Study:
- To enhance the accuracy of predicting adherence lapses in cognitive training for older adults.
- To develop personalized adherence support systems for improved engagement.
Main Methods:
- Utilized data from two prior cognitive training studies.
- Employed deep convolutional neural networks for pattern recognition and prediction.
- Applied domain adaptation (DA) to overcome limited individual training data by leveraging similar user patterns.
- Transformed time series data into image format using Gramian angular fields for participant clustering within DA.
Main Results:
- Deep neural networks and DA demonstrated significant potential in predicting adherence lapses.
- Domain adaptation consistently yielded the highest accuracy across three studies and two independent datasets.
Conclusions:
- Deep learning and DA techniques can support the creation of adherence systems for cognitive training and other health interventions.
- These methods can boost engagement and maximize intervention benefits, enhancing life quality for those at risk of cognitive decline.
- Informs the development of effective interventions to improve aging-related conditions, benefiting individuals and society.
