Radial VIBE Sequence

Motion Robust MRI with Radial VIBE

Radial VIBE is a novel 3D gradient-echo sequence that uses a radial stack-of-stars sampling scheme to acquire the k-space information.
Although radial k-space sampling is known since the early days of MRI and goes back to the seminal work of Paul Lauterbur from 1973, it did not find wide application so far and was never used for routine clinical applications. A key reason is that radial k-space sampling has significantly higher technical requirements, especially regarding the timing accuracy of the gradient fields, as compared to the conventionally employed Cartesian line-by-line sampling or "phase-encoding" sampling. However, with the recent generation of MRI systems, radial k-space sampling is now technically feasible and can be used for routine clinical applications.

Radial sampling offers several promising advantages for clinical scanning:

  • Significantly higher robustness to motion: Radial scans are not affected by the typical MRI ghosting artifacts if the object moves during the scans. Instead, only slight streak artifacts appear (if at all) that usually do not hide pathology.
  • No aliasing artifacts: Because the radial geometry enables to use readout oversampling along the read- and phase-axis without increasing the scan duration, no aliasing effects (infolding) occur if the FOV is selected too small for the object.
  • Continuous update of k-space center: The center of k-space is acquired continuously throughout the scan while each sampled line contains equally important information. Such balanced sampling of k-space makes the acquisition robust to motion and opens up interesting options for motion compensation and for dynamic imaging applications (see GRASP project).
  • Benign undersampling behavior: Instead of creating aliasing effects as seen for Cartesian trajectories when skipping lines, radial undersampling (i.e. a reduction of the number of spokes) creates streaking artifacts in the images, which rather appear as texture patterns and keep the object visible up to even relatively high undersampling factors. Consequently, radial sampling proved to be very useful in combination with undersampling-based techniques for scan acceleration like compressed sensing.

Over the past 2 years, we developed various clinical protocols for this new sequence, focusing on applications where motion is difficult to avoid in routine scanning. These applications include abdominal exams pre and post contrast injection, which are conventionally performed during a breath hold of the patient. Because breath holding poses a big problems for elderly, sick, and pediatric patients, significantly improved image quality can be obtained for these patient populations. Likewise, much better image quality can be obtained in pediatric patients where breath holding is not possible in most cases. Other applications include body areas with frequent motion due to swallowing or eye motion, as well as body areas with completely unavoidable motion like bowel movement in the pelvis.

At time of writing, around 6000 patient scans have been performed at NYU under approval from our institutional review board (IRB) for technical developments to evaluate the performance of the sequence (images have not been used for diagnostic reading). Investigated applications include:

  • Abdominal imaging of incompliant patients / high resolution
  • High resolution abdominal imaging
  • Pediatric imaging
  • Imaging of the head and neck
  • Imaging of the orbits and IAC
  • Whole-body imaging during simultaneous MR-PET exams
  • Breast imaging
  • Lung and chest imaging
  • Enterographies
  • Spine imaging
  • Prostate imaging

The results obtained in a random patient cohort showed that the Radial VIBE sequence works reliable and offers an image contrast similar to that of conventional GRE (VIBE) or MPRAGE. The significantly reduced reduced sensitivity to motion comes at the expense of higher scan duration due to a less efficient coverage of k-space. Thus, Radial VIBE can be the sequence of choice for applications where T1-weighted contrast is required and images are frequently corrupted by motion or pulsation artifacts.

More information about the sequence, in particular about its specific imaging properties, is given in the following presentation:

Principal Investigator: 
Kai Tobias Block
Key Personnel: 
Christian Geppert, Christopher Glielmi, Hersh Chandarana, Sungheon Kim, Ricardo Otazo, Li Feng, Girish, Fatterpekar, Mari Hagiwara, Andrew Rosenkrantz, Sarah Milla

Sponsors

Philanthropic Support

We gratefully acknowledge generous support for radiology research at NYU Langone Medical Center from:
 
• The Big George Foundation
• Raymond and Beverly Sackler
• Bernard and Irene Schwartz

Go to top