Visualizing prostate cancer Image-guided diagnostic strategies

The diagnostic pathway for prostate cancer (PCa) is rapidly evolving, with MRI playing a crucial role in patient selection for screening. However, limitations such as cost and accessibility hinder its widespread use. This thesis addresses these challenges by proposing optimized biopsy strategies and investigating 3D multiparametric ultrasound (3D mpUS) as a cost-effective alternative imaging modality.
Traditional prostate biopsy involves extensive sampling, but a more focused perilesional biopsy approach, targeting MRI-identified lesions, has proven effective while reducing patient burden. Current guidelines recommend combining MRI-targeted and systematic biopsies, but a less invasive approach may be equally effective. Additionally, in patients with negative MRI results, omitting biopsies in certain cases can reduce unnecessary biopsy without increasing mortality.
This thesis also explores ultrasound-based computer-aided diagnosis (CAD) as an alternative to MRI for PCa detection. A study involving 715 patients will use 3D mpUS, incorporating B-mode, shear-wave elastography, and contrast-enhanced ultrasound. The data will help develop an AI algorithm for PCa detection. High-quality pathology-based ground truth data ensures accurate AI training, and a novel framework for mapping pathology to ultrasound images has achieved promising results with a mean registration error of 3.5 mm.
The development of 3D mpUS could revolutionize PCa diagnostics by providing a cost-effective, time-efficient, and accurate alternative to MRI. If validated, it may replace or complement MRI, improving risk stratification while reducing unnecessary biopsies and patient burden.