Clinical evaluation of dispersion MRI for prostate cancer localization: a multicenter study

Multiparametric magnetic resonance imaging (mpMRI) has emerged as a promising tool for prostate cancer (PCa) localization. Due to the complex acquisition procedure and analysis, along with a questioned diagnostic contribution, the inclusion of DCE-MRI in the updated Prostate Imaging and Reporting Data System (PI-RADS2) has been limited to qualitative assessment only, with a marginal weight on the final score. However, lower diagnostic performance has been reported for PI-RADS2 compared to original PI-RADS, evidencing the need to reconsider the role of DCE-MRI. Recently, magnetic resonance dispersion imaging (MRDI) has shown promise for quantitative DCE-MRI of the prostate. In MRDI, the adopted pharmacokinetic (PK) model includes description of the contrast agent intravascular transport, enabling additional assessment of the microvascular architecture. Compared to standard PK analysis in MRI, separate estimation of an arterial input function is no longer necessary, facilitating the clinical routine. Here we evaluate the diagnostic performance of MRDI for PCa localization compared to standard PK analysis by the Tofts model (TM) in a retrospective study including 80 patients from three different Dutch institutions.