Activity-based sleep-wake identification: an empirical test of methodological issues

Area of Science:

• Sleep science
• Biomedical engineering
• Wearable technology

Background:

• Actigraphy is a non-invasive method for sleep-wake monitoring.
• Accurate sleep-wake scoring is crucial for clinical and research applications.
• Understanding the impact of device placement on actigraphy accuracy is essential.

Purpose of the Study:

• To evaluate how actigraph placement (dominant vs. nondominant wrist) affects automatic sleep-wake scoring.
• To assess the robustness of sleep-wake scoring algorithms across different activity levels.
• To explore the utility of twin-wrist actigraphy for artifact detection.

Main Methods:

• Concomitant polysomnography and wrist actigraphy data were collected from adults and adolescents.
• Actigraphy data were gathered from both dominant and nondominant wrists.
• Sleep-wake scoring algorithms were developed and validated against polysomnography.

Main Results:

• Actigraphy placement did not significantly alter sleep-wake scoring algorithm performance.
• Agreement rates with polysomnography ranged from 91-93% for both wrists.
• The scoring algorithm demonstrated robustness to moderate changes in activity levels.
• Twin-wrist actigraphy facilitated the identification of motion artifacts.

Conclusions:

• Actigraphy device placement on either wrist provides comparable sleep-wake scoring accuracy.
• The developed scoring algorithm is reliable and robust for sleep-wake assessment.
• Twin-wrist actigraphy offers an advantage in detecting potential artifacts, improving data quality.