Application of CT imaging parameters in predicting patellofemoral joint instability: a multi-parameter combined analysis

Patellofemoral joint instability (PFI) is a common knee disorder characterized by abnormal patellar tracking and recurrent instability, which may lead to functional impairment and progressive patellofemoral joint degeneration. Accurate diagnosis of PFI remains challenging due to the multifactorial nature of its pathogenesis. Computed tomography (CT)-based measurements provide important quantitative information regarding lower limb alignment and rotational abnormalities; however, the diagnostic value of multiple CT parameters and their combined predictive performance require further investigation. This study aimed to evaluate the diagnostic value of these CT parameters individually and in combination for PFI. This study recruited 70 patients diagnosed with patellofemoral joint instability, along with 45 asymptomatic controls. Knee CT data were collected, and TT-TG, TT-ME, KJRA, and TT-TA were measured using standardized methods. Interobserver reliability was assessed for TT-ME. Univariable and multivariable regression analyses were conducted using SPSS 19.0 to develop the multi-parameter predictive model. The model’s diagnostic performance was evaluated with ROC curve analysis, while calibration curves and decision curve analysis (DCA) were used to confirm its clinical applicability. The patient group exhibited significantly higher values in tibial tubercle torsion angle (TT-TA), knee joint rotation angle (KJRA), tibial tubercle-midepicondylar distance (TT-ME), and tibial tubercle-trochlear groove distance (TT-TG) compared to controls ( P < 0.05). In contrast, patients with patellofemoral instability were younger than controls ( P < 0.05). Multivariable analysis revealed TT-ME (OR = 1.535, P < 0.05), TT-TA (OR = 1.131, P < 0.05), and age (OR = 0.914, P < 0.05) as independent predictors of patellofemoral joint instability. The multi-parameter predictive model demonstrated superior diagnostic accuracy (AUC = 0.929), outperforming individual parameters (TT-ME AUC = 0.893; TT-TA AUC = 0.722). Decision curve analysis further suggested potential clinical utility, showing consistent net benefit gains across a wide probability threshold spectrum. TT-ME distance, TT-TA, and age contributed to the diagnostic model for patellofemoral joint instability. Their combined application significantly improves the diagnostic accuracy of CT imaging. Compared to individual parameters, the multi-parameter predictive model exhibits superior overall diagnostic accuracy, highlighting its valuable clinical utility.