Mandible behaviour interpretation during wakefulness, sleep and sleep-disordered breathing

Area of Science:

• Sleep Medicine
• Respiratory Physiology
• Biomedical Signal Analysis

Background:

• Mandible movement (MM) signals offer insights into patient activity during sleep.
• Visual interpretation of MM can aid in assessing sleep-wake states and detecting respiratory events.
• Standardization of MM analysis is crucial for reliable clinical application.

Purpose of the Study:

• To train independent scorers to accurately recognize specific mandible movement (MM) signal patterns.
• To evaluate the intrascorer reproducibility of MM signal interpretation.
• To assess the interscorer variability in diagnosing sleep-wake states and respiratory events from MM signals.

Main Methods:

• Collected MM signals in the mid-sagittal plane from 40 patients.
• Classified MM signals into seven distinct patterns: active wakefulness (AW), quiet wakefulness/sleep (QW/S), sleep snoring (SS), obstructive apnea (OAH), mixed apnea (MA), respiratory related arousal (RERA), and central apnea (CAH).
• Trained four scorers and assessed their diagnostic accuracy across two reading sessions using Cohen's kappa for agreement.

Main Results:

• Intrascorer reproducibility ranged from 0.68 to 0.88, indicating good consistency.
• Interscorer agreement was 0.68 and 0.74 across sessions, demonstrating acceptable variability.
• Overall scorer accuracy was 75.2%, with central apnea events (CAH) being the most challenging (65.6% accuracy). Recognition of abnormal respiratory events (OAH, CAH, MA, RERA) was excellent (90.7% interscorer agreement).

Conclusions:

• Visual analysis of isolated mandible movement (MM) signals is a viable method for diagnosing sleep-wake states.
• MM signal interpretation can successfully identify normal and abnormal respiration, including respiratory effort.
• While challenging, specific patterns like central apnea events can be discerned with further training and refinement.