Retrospective Validation of a CAD System Using Multiparametric Transrectal Ultrasound for Clinically Significant Prostate Cancer Localization

This study evaluates a computer-aided diagnosis (CAD) system using 3D multiparametric transrectal ultrasound (mpUS) for detecting clinically significant prostate cancer (ASPCA). The CAD system was trained with images from patients undergoing radical prostatectomy and those with negative MRI or biopsies. In a retrospective validation of 220 patients, the CAD system demonstrated a sensitivity of 81.2% and specificity of 39.3% at the lesion level, with a positive predictive value (PPV) of 77.3% and a negative predictive value (NPV) of 45.1%. For Gleason Group≥3 csPCa, the sensitivity increased to 88.4%. While the CAD shows promise comparable to MRI, further prospective validation is necessary.