Performance Evaluation of a Smart Bed Technology against Polysomnography.

Area of Science:

• Biomedical Engineering
• Sleep Science
• Wearable Technology

Background:

• Smart beds integrate advanced sensors for non-invasive physiological monitoring.
• Ballistocardiography (BCG) combined with machine learning shows potential for sleep tracking.

Purpose of the Study:

• To evaluate the accuracy of the Sleep Number smart bed in measuring heart rate (HR), breathing rate (BR), and sleep states against polysomnography (PSG).
• To assess the smart bed's performance in estimating epoch-by-epoch and summary sleep variables.

Main Methods:

• Forty-five participants underwent one night of sleep monitoring using both the Sleep Number smart bed and standard PSG.
• Data analysis included Bland-Altman analysis, Pearson correlation for HR and BR, and Cohen's kappa, ROC analysis for sleep/wake classification.

Main Results:

• High correlations were found between smart bed and PSG for epoch-by-epoch HR (r=0.81) and BR (r=0.71).
• Mean overnight HR and BR estimations also showed strong agreement (r=0.94 for HR, r=0.96 for BR).
• Sleep vs. wake detection demonstrated good performance with an accuracy of 0.86 and precision of 0.90.

Conclusions:

• The Sleep Number smart bed provides reliable, unobtrusive metrics for characterizing human sleep.
• The technology shows significant potential for real-world sleep research and monitoring applications.