Development and Validation of an Ultrasound-Based Clinical Radiomics Nomogram for Diagnosing Gouty Arthritis

This study aimed to develop and validate a diagnostic model for gouty arthritis by integrating ultrasonographic radiomic features with clinical parameters.

A total of 604 patients suspected of having gouty arthritis were enrolled and randomly divided into a training set (n = 483) and a validation set (n = 121) in a 4:1 ratio. Univariate and multivariate analyses were conducted on the clinical data to identify statistically significant clinical features for constructing an initial diagnostic model. Key radiomic features were identified in the training set using least absolute shrinkage and selection operator (LASSO) regression analysis to establish a radiomic model. A composite clinicoradiomic nomogram was then developed by combining clinical (such as C-reactive protein, erythrocyte sedimentation rate and uric acid level) and radiomic features through logistic regression. The predictive performance of the clinical model, radiomic model and clinicoradiomic nomogram was evaluated in the validation set using receiver operating characteristic curves, calibration curves and decision curve analysis.

The clinicoradiomic nomogram, which integrated imaging features and clinical characteristics via logistic regression, demonstrated superior predictive performance in the validation set, with an area under the curve (AUC) of 0.936 (95% CI: 0.885-0.986), surpassing both clinical (AUC = 0.924; 95% CI: 0.873-0.976) and radiomic models (AUC = 0.828; 95% CI: 0.738-0.918) alone. Decision curve analysis further confirmed the clinical utility of this model, particularly in differentiating between gouty and non-gouty arthritis.

Compared with standalone clinical or radiomic models, the ultrasonography-based clinicoradiomic model exhibited enhanced predictive accuracy for diagnosing gouty arthritis, presenting a novel and promising approach for the early diagnosis and management of gouty arthritis.