A radiomics model based on transrectal ultrasound for predicting prostate cancer.

Abstract

Prostate cancer (PCa) is one of the most common neoplasms in men. However, the value of ultrasound-based radiomicsfor diagnosing PCa remains uncertain. We retrospectively analyzed ultrasonic and clinicaldata from 373 patients. Patients were divided into two groups according to the pathological results. Radiomics features wereextracted from TRUS, and we screened the optimal features to construct radiomics models. Relationships between clinicalcharacteristics and prostate lesions were identified by univariate and multivariate logistic regression analysis. Finally, aclinical-radiomics model was developed, and then visualized in the form of a nomogram. Of the 373 patients, 178had benign disease and 195 had malignant disease. The support vector machine (SVM) classification model showed the bestperformance, while the diagnostic performance of the clinical model was poorer than that of the radiomics model (p<0.05) orthe combined (clinical-radiomics) model (p<0.05). In general, the combined model demonstrated the highest AUC and provedto be more advantageous. The prediction model we constructed based on TRUS predicted PCa preoperativelywith high efficiency. In addition, combining radiomics with clinical factors improved diagnostic accuracy.