Using Smart Homes to Detect and Analyze Health Events

Area of Science:

• Gerontology
• Health Informatics
• Data Mining

Background:

• Instrumented smart homes enable unobtrusive monitoring of behavior in natural settings.
• Understanding behavior-health relationships is crucial for managing health conditions.

Purpose of the Study:

• To introduce a behavior change detection (BCD) approach for identifying behavioral shifts associated with health events.
• To analyze behavior changes in older adults living in smart homes.

Main Methods:

• Collected sensor data from smart homes and utilized activity recognition.
• Applied behavior change detection (BCD) to quantify and analyze changes in activity timings and durations.
• Evaluated the approach using three case studies of older adults experiencing major health events.

Main Results:

• Demonstrated that significant behavior changes, consistent with medical literature, occur during major health events.
• Successfully and automatically detected these behavior changes using the BCD approach.
• Validated the effectiveness of BCD in analyzing behavioral effects of health conditions.

Conclusions:

• The proposed smart home, activity recognition, and BCD algorithms are effective data mining techniques.
• This approach provides valuable insights into the behavioral impacts of health conditions.
• Automated detection of behavior changes can aid in understanding and managing health in older adults.