FastMRI in Prostate Cancer
With a recent change in national and European guidelines for diagnosing suspected prostate cancer (PCa), multiparametric MRI has been recommended as the first-line diagnostic workup. Moreover, MRI is expected to contribute intensively to an active surveillance strategy. As expected, this increases the number of MRI scans and costs.
This PhD project aims to develop faster MRI sequences. We intend to develop and implement MRI sequences that under-sample k-space for accelerated AI-based reconstruction. Speeding up MRI protocols will decrease the expensive MRI in-bore time, consequently decreasing healthcare costs. We also aim at developing and implementing MRI sequences that steer acquisition towards a region of interest by rotating imaging planes or by locally increasing imaging resolution. Depending on the reconstruction strategies developed, we will develop reduced k-space sequences to optimize AI reconstruction and detection. This will pave the way for AI techniques for MRI diagnosis and intervention of PCa developed at UMCG and RadboudUMC.
The ultimate goal is to integrate AI in MRI for faster diagnosis and improved therapeutic applications for PCa and, eventually, future expansion to other (clinical) MRI applications.
(in collaboration with RUMC and UMCG),
from (PDF) fastMRI: An Open Dataset and Benchmarks for Accelerated MRI (researchgate.net), Using MRI for Prostate Cancer Diagnosis Equals or Beats Current Standard | Imaging Technology News (itnonline.com)