Background/Objectives: The therapeutic effectiveness of chronic rhinosinusitis with nasal polyposis (CRSwNP) depends on an accurate diagnosis that identifies disease characteristics, evaluates sinus patency, and detects paranasal sinus obliteration. This study aims to assess a novel artificial intelligence (AI) system integrated with radiomic analysis for the radiological evaluation of CRSwNP, developing a reliable and predictive clinical-radiological scoring system. Methods: This study retrospectively evaluates CT scans of patients with CRSwNP. Image analysis was performed using Radiomica LifeX (Local Image Features Extraction) version 7.5. The extracted densitometric volumes were compared to the Lund-Mackay Score (LMS) to develop a novel scoring system (P-ABCD score) and assess its radiomic predictive capability. Results: Twenty patients with CRSwNP undergoing Dupilumab therapy participated in this study. The P-ABCD score, derived from sinus CT imaging data, served as a valuable objective measure of clinical improvement following CRSwNP treatment. Conclusions: Advanced radiomic imaging techniques of the sinus cavity provide precise volumetric data combined with texture analysis. These techniques offer high sensitivity by accurately quantifying the true extent of inflammatory involvement in the paranasal sinuses, enabling effective disease stratification.
Masieri et al. (Thu,) studied this question.