Objective: To develop a risk prediction model for prostate cancer (PCa) by integrating Shear Wave Elastography (SWE) with Multiparametric Magnetic Resonance Imaging (mpMRI), thereby improving screening accuracy and specificity while reducing unnecessary invasive procedures.
Methods: A total of 479 patients who visited Sun Yat-sen Memorial Hospital between May 2019 and July 2023 were included in this retrospective study, with 162 diagnosed with PCa. The patients were randomly divided into a training set (349 cases) and a validation set (130 cases). The primary measurements consisted of the Young's modulus from SWE, the PI-RADS score from mpMRI, and laboratory indicators such as total PSA (tPSA), free PSA (fPSA), and their densities. A multifactorial prediction model integrating imaging and clinical data was constructed and validated.
Results: The combined model incorporating SWE and mpMRI exhibited high accuracy and robustness in diagnosing PCa, with area under the curve (AUC) values of 0.92 for the training set and 0.91 for the validation set, significantly outperforming individual indicators (P<0.001). The model achieved a sensitivity of 94.87% and a specificity of 96.12%, indicating superior performance in distinguishing PCa from benign lesions. Receiver operating characteristic (ROC) curve analysis and DeLong's test confirmed that the combined model exhibited the highest diagnostic accuracy, reducing false positives and minimizing unnecessary biopsies.
Conclusions: The multifactorial prediction model integrating both imaging and clinical data provides a more precise and reliable tool for the early diagnosis of PCa, with significant potential for clinical application.
Keywords: Shear wave elastography; multiparametric MRI; prediction model; prostate cancer.
AJCR Copyright © 2025.