We constructed a morphological model of diffusion in the prostate from a limited number of diffusion-weighted images to increase the sensitivity of such diffusion imaging to the presence of prostate cancer. Estimating the measurement error (9.9%) and characterizing the prostate from a large public dataset (n=206) has shown morphological relationships (|r|>0.5) and provided distributions and relationships within the available ADC measures. A model can then be used to give expected values to test against, and enable much larger datasets to be synthesized with the aim of testing various machine learning approaches.
This abstract and the presentation materials are available to members only; a login is required.