3D Camera and Pulse Oximeter for Respiratory Events Detection

Area of Science:

• Medical Devices
• Sleep Medicine
• Biomedical Engineering

Background:

• Traditional polysomnography (PSG) for sleep apnea detection relies on sensors that can be uncomfortable and prone to artifacts.
• Accurate detection of respiratory events is crucial for diagnosing and managing sleep-related breathing disorders.

Purpose of the Study:

• To develop a respiratory movement signal from a 3D time-of-flight camera.
• To evaluate the combination of this signal with SpO2 for detecting respiratory events comparable to PSG.

Main Methods:

• A novel algorithm was developed to detect reduced respiratory movement and SpO2 desaturation.
• The method was validated on 61 patients using synchronized 3D video and PSG recordings.
• Performance was assessed using apnea-hypopnea index (AHI) correlation, severity classification, and epoch-based analysis (accuracy, sensitivity, specificity, Cohen's kappa).

Main Results:

• The predicted AHI showed excellent agreement with manual PSG (ICC r = 0.94).
• Severity classification achieved 80% accuracy.
• Epoch-based analysis yielded Cohen's kappa = 0.72, accuracy = 0.88, sensitivity = 0.80, and specificity = 0.91.

Conclusions:

• The 3D camera and SpO2-based detection method demonstrates excellent reliability and substantial agreement with PSG.
• This non-contact approach offers a potential alternative to traditional, potentially uncomfortable, respiratory monitoring sensors.