Grant: Project Grant

Prostate cancer (PCa) is the most common cancer diagnosed in Canadian men, with about 25,000 new cases per year. The gold standard for PCa diagnosis and prognosis is ultrasound (US)-guided needle biopsy with pathologic grading. The biopsy procedure is not patient specific; rather, it involves taking "blind" biopsy cores, leading to frequent re-biopsy or over diagnosis of insignificant cancer. There is an acute need for patient-specific targeting techniques that have higher chance of detecting early aggressive PCa. Aggressive PCa must be diagnosed early and accurately to appropriately treat patients at higher risk of morbidity and mortality. Meanwhile, it is critical to balance against over-treatment of indolent PCa, of which is more appropriately managed using an active surveillance approach. The best biopsy technique today is to localize PCa with magnetic resonance imaging (MRI), and use its fusion with US to improve the detection rate of aggressive PCa. However, this method alone is not able to accurately differentiate indolent PCa and cannot reliably detect small lesions of aggressive PCa. In this project, we propose a turnkey technology, where subtle variations of backscattered US signals from a tissue location over a short time span, i.e., temporal enhanced US, are used to detect physical properties of the underlying tissue using a machine-learning framework. Our objective is to establish the efficacy of this technology in identifying clinically significant PCa in real-time US, so that accurate biopsy targeting is achieved. In a series of clinical studies, we will develop and deploy this technology across multiple institutions. This technology is expected to positively and rapidly benefit the lives of all patients suffering from PCa. Collaboration with leading investigators in Canada and USA will create an effective environment for leading international development of next generation cancer diagnosis imaging tools.