Machine learning model uses urinary extracellular vesicle physical parameters to improve prostate cancer diagnosis, reducing need for invasive biopsies.
