Meeting Banner
Abstract #0583

High Resolution Dynamic Susceptibility Contrast Perfusion Imaging Using Multi-Echo Spirals and Temporal Compressed Sensing-Based Reconstruction

Julia V. Velikina1, Youngkyoo Jung2, Aaron A. Field3, Alexey A. Samsonov3

1Medical Physics, University of Wisconsin - Madison, Madison, WI, United States; 2Radiology, Wake Forest University, Winston-Salem, NC, United States; 3Radiology, University of Wisconsin - Madison, Madison, WI, United States

We propose a novel approach to dynamic contrast susceptibility perfusion-weighted imaging using a combination of multi-echo spiral acquisition and compressed sensing-type regularized reconstruction based on 2nd derivative in temporal dimension to overcome limitations of standard EPI-based perfusion imaging technique. The proposed technique allows for a significant increase of in-plane spatial resolution and improved whole-head coverage (voxel size 1.375x1.375x5 mm) without compromising temporal resolution (1.35 ms). Availability of four different echo times allows for unbiased R2* mapping and more accurate/artifact-free estimation of relevant haemodynamic perfusion parameters (cerebral blood volume/flow and mean transit time).

Keywords

acceleration accompanied according achieved acquisition adding advanced aliasing alone anatomical angiography approaches arterial artifacts attainable avoid bias brain cerebral clear clinically coil combination compressed compromising concentration context contrast course coverage curve decay decreases degree dependencies derived detail detect developed dictated differentiable discovery distortions done dotted double dynamic efficient efficiently either enabling encoding enumerated estimation exams exploiting frame frames geometric gray head highly human illustrate improved improvement individual injected input institution inter limited loss mapping maps measured measuring minimized model models navigation necessitating noise nominal nonlinear norm note novel often operator package peak performance perfusion pixel previously processed promoting proposed quadratic quantification quantitative radiology rapid rationale readout receivers reconstructed reconstruction regularization relapsing relevant residual resolution scanner segmented self sense sensing separate series shorter shot slices smoothly software source space sparsity spatial spiral spirals stroke subjects sufficient support suppresses susceptibility systems target temporal traditionally transform treatment tumor type typically utilizes vessels volume white