The predictive value of photogrammetry for obstructive sleep apnea.

Area of Science:

• Medical imaging
• Sleep medicine
• Diagnostic modeling

Background:

• Obstructive sleep apnea (OSA) diagnosis relies on polysomnography, often preceded by physical examination.
• Predictive models for OSA can aid in early identification and management.
• Objective measurements of upper airway structures are needed for improved OSA prediction.

Purpose of the Study:

• To develop and validate a prediction model for obstructive sleep apnea (OSA) using photographic measurements of upper airway structures.
• To compare the predictive performance of a photogrammetry-based model against a model using general physical examination findings.
• To assess the combined predictive power of photographic and physical examination measurements for OSA.

Main Methods:

• Logistic regression analysis was employed to construct prediction models.
• Participants with suspected OSA underwent physical examination and oropharyngeal photography before polysomnography.
• Data from 197 eligible participants were analyzed, with 74% diagnosed with OSA.

Main Results:

• A photogrammetry model using four measurements (tongue area, uvula area, frenulum length, retroposition distance) achieved 82.7% correct classification.
• The photogrammetry model demonstrated superior predictive performance (AUC=0.90) compared to the physical examination model (AUC=0.80).
• Combining photographic and physical measurements further improved prediction accuracy (AUC=0.93).

Conclusions:

• Photogrammetry provides detailed data on upper airway abnormalities associated with OSA.
• Prediction models incorporating photographic measurements show significant utility in diagnosing OSA.
• Integrating photographic and physical examination data enhances OSA prediction efficacy.