Encrypted login | home

Program Information

T2-Weighted 4D-MRI Using An Isotropic 3D Data Acquisition and Retrospective K-Space Binning

no image available
Z Deng

Z Deng*, A Christodoulou , W Yang , X Bi , J Shaw , B Fraass , R Tuli , D Li , Z Fan , Cedars-Sinai Medical Center, Los Angeles, CA

Presentations

TU-D-601-6 (Tuesday, August 1, 2017) 11:00 AM - 12:15 PM Room: 601


Purpose: 4D-MRI is an emerging technique for characterizing the motion of body organs in radiotherapy planning. T2 contrast weighting has proved valuable for tumor visualization. Existing T2-weighted 4D-MRI methods rely on resorting multi-slice T2-weighted 2D images, which is associated with potential stitching artifacts and limited temporal and slice resolution. To overcome these limitations, we propose a T2-weighted 4D-MRI technique using a 3D acquisition and self-gating-based retrospective k-space binning.

Methods: The proposed method is based on a 3D hybrid-radial gradient recalled echo (GRE) sequence. During each period of one second, a T2 preparation (T2prep) module (~40ms) is followed by 17 cycles of one self-gating and 9 imaging lines. K-space is filled using golden angle ordering and Gaussian-density randomized partition ordering. After temporal binning (20 phases) of k-space data, each phase set is reconstructed using a low-rank tensor imaging method accounting for T1 recovery. The 4D MRI technique was tested in 3 volunteers on a 3-Tesla system. The sequence continuously ran for 10 minutes with a coronal field-of-view of 384x384x128mm3 and isotropic 2mm resolution. 2D real-time (382ms/frame) images were also acquired to provide the reference for motion evaluation.

Results: Initial T2-weighting diminished quickly due to T1 signal recovery. Our proposed method effectively eliminated T1-weighting, allowing the combination of imaging lines acquired at different portions of the recovery period to generate T2-weighted images. Compared with using only the imaging lines immediately following the T2prep, this strategy used data from across the recovery period to yield much better image quality without affecting image contrast. The 4D MRI technique was in good accordance with 2D real-time MRI in terms of motion amplitudes of the liver.

Conclusion: This work presents a novel T2-weighted 4D-MRI technique with isotropic resolution and reasonable image quality. Further technical improvements are underway to make the technique more clinically useful.


Contact Email: