Background
Prostate cancer detection for decades was done using systematic ultrasound biopsies in lack of a modality that can detect presence and location of a cancer. The last decade MRI has evolved into a modality that can succesfully detect and localize prostate cancer. This allows to avoid biopsies in case no cancer is present or to guide the biopsy to reduce increase certainty about aggressiveness. Guided biopsies are performed in the MRI itself, but mostly using ultrasound fusion. Fusion is easier for the patient and more cost-effective for healthcare, but slightly less effective than MRI biospy.
Challenges: * Prostate MRI reading performance depends very strongly on expertise, * Fusion biopsy misses smaller and less aggressive lesions.
Aim
Develop AI to improve reading prostate MRI and fusion biopsy.
