Modern diagnostics: ultrasound elastography and magnetic resonance imaging in initial evaluation of testicular cancer

Abstract

Objectives: Differentiating benign from malignant testicular lesions is essential to avoid unnecessary surgery and ensure timely intervention. While conventional ultrasound remains the first-line imaging method, elastography and MRI provide additional functional and structural information. This study assesses the diagnostic utility of testicular elastography and magnetic resonance imaging (MRI) in differentiating benign and malignant testicular lesions. Methods: Patients with sonographically detected testicular masses were retrospectively evaluated using elastography, scrotal MRI, and tumor markers. Quantitative and qualitative imaging findings, lesion size, and laboratory values were recorded. Statistical analyses included Fisher's exact test, logistic regression, Receiver operating characteristic (ROC) analysis, and machine learning classification using the eXtreme Gradient Boosting (XGBoost) model. Results: Our analysis demonstrated that quantitative elastography significantly predicts malignancy (p = 0.042), with high diagnostic power (Area under the Curve [AUC] = 0.91). Additionally, Doppler ultrasound vascularity showed a statistically significant positive correlation with malignancy risk (p = 0.0033), highlighting its value in diagnosis. In contrast, qualitative elastography and MRI contrast enhancement lacked statistical significance. MRI diffusion restriction provided valuable insight into malignancy risk, though the latter did not reach statistical significance. An XGBoost model incorporating imaging and laboratory parameters showed high precision in detecting malignancies (85.7%). Conclusion: Findings underscore the importance of Doppler ultrasound as the primary diagnostic tool, with elastography and diffusion MRI aiding in cases of diagnostic ambiguity. The combined use of all imaging modalities and laboratory markers enhances diagnostic accuracy in evaluating testicular masses.