GRAPPA reconstructed wave-CAIPI MP-RAGE at 7 Tesla

Jolanda M Schwarz; Eberhard D Pracht; Daniel Brenner; Martin Reuter; Tony Stöcker

doi:10.1002/mrm.27215

GRAPPA reconstructed wave-CAIPI MP-RAGE at 7 Tesla

Magn Reson Med. 2018 Dec;80(6):2427-2438. doi: 10.1002/mrm.27215. Epub 2018 Apr 16.

Authors

Jolanda M Schwarz¹, Eberhard D Pracht¹, Daniel Brenner¹, Martin Reuter^{1

2

3}, Tony Stöcker^{1

4}

Affiliations

¹ German Center for Neurodegenerative Diseases (DZNE), Bonn, Germany.
² Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital, Harvard Medical School, Charlestown, Massachusetts.
³ Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Cambridge, Massachusetts.
⁴ Department of Physics and Astronomy, University of Bonn, Germany.

PMID: 29663507
DOI: 10.1002/mrm.27215

Abstract

Purpose: The aim of this project was to develop a GRAPPA-based reconstruction for wave-CAIPI data. Wave-CAIPI fully exploits the 3D coil sensitivity variations by combining corkscrew k-space trajectories with CAIPIRINHA sampling. It reduces artifacts and limits reconstruction induced spatially varying noise enhancement. The GRAPPA-based wave-CAIPI method is robust and does not depend on the accuracy of coil sensitivity estimations.

Methods: We developed a GRAPPA-based, noniterative wave-CAIPI reconstruction algorithm utilizing multiple GRAPPA kernels. For data acquisition, we implemented a fast 3D magnetization-prepared rapid gradient-echo wave-CAIPI sequence tailored for ultra-high field application. The imaging results were evaluated by comparing the g-factor and the root mean square error to Cartesian CAIPIRINHA acquisitions. Additionally, to assess the performance of subcortical segmentations (calculated by FreeSurfer), the data were analyzed across five subjects.

Results: Sixteen-fold accelerated whole brain magnetization-prepared rapid gradient-echo data (1 mm isotropic resolution) were acquired in 40 seconds at 7T. A clear improvement in image quality compared to Cartesian CAIPIRINHA sampling was observed. For the chosen imaging protocol, the results of 16-fold accelerated wave-CAIPI acquisitions were comparable to results of 12-fold accelerated Cartesian CAIPIRINHA. In comparison to the originally proposed SENSitivity Encoding reconstruction of Wave-CAIPI data, the GRAPPA approach provided similar image quality.

Conclusion: High-quality, wave-CAIPI magnetization-prepared rapid gradient-echo images can be reconstructed by means of a GRAPPA-based reconstruction algorithm. Even for high acceleration factors, the noniterative reconstruction is robust and does not require coil sensitivity estimations. By altering the aliasing pattern, ultra-fast whole-brain structural imaging becomes feasible.

Keywords: CAIPIRINHA; GRAPPA; MP-RAGE; parallel imaging; wave-CAIPI.

MeSH terms

Algorithms
Artifacts
Brain / diagnostic imaging*
Humans
Image Enhancement / methods
Image Interpretation, Computer-Assisted / methods
Image Processing, Computer-Assisted / methods*
Imaging, Three-Dimensional
Magnetic Resonance Imaging*
Neuroimaging / methods*
Software