Correlation of Patient-Reported Symptoms With Rhinogram Features Beyond Simple Airway Resistance.

Area of Science:

• Rhinology and Otolaryngology
• Medical Engineering
• Data Science in Medicine

Background:

• Rhinomanometry is a standard for nasal airway assessment but often shows weak correlation with patient symptoms.
• Rhinomanometry curves (rhinograms) contain complex data beyond simple nasal resistance at 150 Pascals (Pa).
• Exploring these curve features may enhance clinical relevance and patient outcome correlation.

Purpose of the Study:

• To investigate the correlation between novel rhinogram curve features and patient-reported outcomes (PROMs).
• To compare the predictive power of curve-derived features against standard nasal airway resistance (NAR) measurements.
• To develop an AI predictive model using significant curve features.

Main Methods:

• A cross-sectional study involving 601 patients from a rhinology clinic.
• PROMs collected included nasal obstruction severity (ordinal scale) and visual analogue scale (VAS).
• Rhinomanometry curves were analyzed using polynomial fitting to extract 835 features; Spearman's rho (ρ) was used for correlation analysis.

Main Results:

• Curve-derived features showed stronger correlations with VAS (ρ=0.305) and ordinal nasal obstruction (ρ=0.230) compared to NAR at 150 Pa (VAS: ρ=0.222; obstruction: ρ=0.112).
• AI models predicted VAS with ρ=0.133 and nasal obstruction with ρ=0.117.
• The study identified significant correlations between detailed rhinogram features and patient-reported nasal obstruction levels.

Conclusions:

• Novel analysis of rhinomanometry curves yields features that correlate better with PROMs than standard NAR.
• While correlations are weak-to-moderate, these curve-derived features offer enhanced clinical information.
• Rhinogram analysis holds potential for more comprehensive nasal airway assessment beyond simple resistance values.