A Pilot Study to Detect Agitation in People Living with Dementia Using Multi-Modal Sensors

Area of Science:

• Gerontology
• Biomedical Engineering
• Neuroscience

Background:

• Behavioral and psychological symptoms, including agitation and aggression, are common in people living with dementia (PLwD).
• Agitation in PLwD causes significant distress and elevates the risk of injury for both patients and caregivers.
• Current methods for detecting agitation may not fully capture the complexity of these symptoms.

Purpose of the Study:

• To present a novel multi-modal wearable device for detecting agitation in PLwD.
• To identify key sensor-derived features most relevant for accurate agitation detection.
• To evaluate the efficacy of combining multi-modal sensor data versus single-sensor data for agitation detection.

Main Methods:

• Utilized a multi-modal wearable device to capture motion and physiological signals from PLwD.
• Extracted and analyzed relevant features from the collected sensor data.
• Collected data from 17 PLwD over 600 days in a Specialized Dementia Unit.

Main Results:

• The study identified specific sensor signal features crucial for detecting agitation.
• Combining data from multiple sensor modalities proved more effective than using data from a single sensor.
• The findings underscore the value of multi-modal data in understanding and detecting agitation.

Conclusions:

• Multi-modal wearable devices offer a promising approach for objective agitation detection in PLwD.
• Feature analysis highlights critical indicators for improved agitation monitoring.
• This pilot study provides a foundation for developing advanced assistive technologies for dementia care.