This prospective clinical study aims to retrain and validate the PCaVision AI algorithm for the detection and localization of clinically significant prostate cancer (csPCa) using multiparametric ultrasound (mpUS) with two different ultrasound contrast agents: Sonovue and Optison. Data will be collected using a commercial version of the LOGIQ E10 ultrasound machine, with full histopathological correlation in patients undergoing radical prostatectomy.
The PCaVision CAD system was initially developed using Sonovue contrast and a research-only ultrasound machine (LOGIQ E10 R3). In this study, the algorithm will be retrained using data acquired from the LOGIQ E10 R4, a commercially available system, and using both Sonovue and Optison as contrast agents. The goal is to enhance generalizability and enable broader clinical adoption. The study includes two cohorts: patients scheduled for radical prostatectomy (for training and voxel-level validation) and patients with a negative prostate MRI (for patient-level specificity evaluation). All participants will undergo 3D transrectal mpUS with both contrast agents in a single session. Full prostate histology from RP specimens will serve as the reference standard.
Study Type
INTERVENTIONAL
Allocation
NA
Purpose
DIAGNOSTIC
Masking
NONE
Enrollment
212
Participants will receive 3D mpUS using both Sonovue and Optison contrast agents during a single transrectal imaging session.
Amsterdam UMC
Amsterdam, Netherlands
Sensitivity and Specificity of PCaVision Algorithm (Sonovue and Optison) Compared to Histopathology
Sensitivity and specificity of the PCaVision algorithm in detecting and localizing clinically significant prostate cancer (csPCa, Gleason Grade Group ≥2) at the voxel level, based on 3D multiparametric ultrasound (mpUS) images using Sonovue and Optison, validated against full prostate histology after radical prostatectomy.
Time frame: From baseline imaging to histopathology report (typically within 2-4 weeks post-surgery)
Specificity of PCaVision in MRI-Negative Cohort
Evaluation of the specificity of the PCaVision algorithm for csPCa detection at the patient level in men with a negative prostate MRI (PI-RADS ≤2), using mpUS imaging with Sonovue and Optison.
Time frame: From baseline imaging to completion of image processing and analysis (within 2 weeks of scan)
Diagnostic Performance Comparison Between Sonovue and Optison
Head-to-head comparison of diagnostic performance (sensitivity, specificity, AUC) of PCaVision on mpUS scans obtained with Sonovue versus Optison, using RP histopathology as reference. Evaluated at voxel level and using simulated biopsy protocols.
Time frame: Imaging and analysis completed within 4 weeks post-surgery
This platform is for informational purposes only and does not constitute medical advice. Always consult a qualified healthcare professional.