4451
The impact of shorter acquisition time in MRF: Long term repeatability and reproducibility study on ISMRM/NIST phantom and volunteers.1Department of Radiological Technology, Nagoya University Hospital, Nagoya, Japan, 2Department of Radiology, Nagoya University Hospital, Nagoya, Japan, 3MR Research & Collaboration, SIEMENS Healthcare K.K., Tokyo, Japan, 4Magnetic Resonance, Siemens Healthcare GmbH, Erlangen, Germany, 5Friedrich-Alexander-Universität Erlangen-Nürnberg, Erlangen, Germany
Synopsis
This study focused on the stability of MRF in a phantom and
volunteers, and explored the feasibility of MRF with a shorter acquisition
time. Phantom scans on 40 days and volunteer scans on 5 days over 3 months showed
comparable repeatability and reproducibility of T1 and T2 values between MRF
with acquisition times of 41 sec and 20 sec. Shorter acquisition time has the
potential to expand the clinical usage of MRF.
Introduction
MR Fingerprinting (MRF) is a novel concept for simultaneous
acquisition of multiple tissue parameters by varying acquisition parameters in
the sequence, and has the potential to provide clinically valuable information
for MR diagnosis. The FISP-MRF sequence applies varying flip angles (FA) and repetition
times (TR), and the number of TRs determines the acquisition time. However,
there have been few reports to systematically evaluate effects with different acquisition
lengths of FISP-MRF. The purpose of
this study was to evaluate the repeatability and reproducibility of FISP-MRF
with different acquisition lengths (long MRF/ short MRF) in a phantom and human volunteers, and to
explore the feasibility of MRF with shorter scan times.Methods
All studies were performed on a MAGNETOM Skyra 3T (Siemens
Healthcare, Erlangen, Germany) with a 20-channel head receiver array. The
ISMRM/NIST system phantom [1], which contains multiple spheres covering a wide
range of T1 and T2 values, was placed in the magnet for more than 30 min before
the acquisition to let the liquid motion settle. The phantom was scanned with a
prototype implementation of a 2D spiral FISP-MRF sequence [2,3]. Two slices,
at the T1 and T2 layers, were scanned on 40 days with MRF using two different
acquisition lengths. One employs 3000 TRs with 41 sec acquisition time per
slice (long MRF), and the other has 1500 TRs with 20 sec acquisition time
(short MRF). Both sequences consist of FA varying from 0° to 74° and TR varying
from 12 to 15 msec. Before FISP-MRF acquisitions, the temperature of the phantom
was measured. A fast RF field (B1+) map [4] was obtained, and a correction
based on the B1+ map was applied for each pixel in combination with a
dictionary extended by a B1+ dimension. T1 and
T2 values of each sphere were obtained from a circular region of interest (ROI)
that was manually drawn on the T1 and T2 map to exclude edge pixels. Note that
we only analyzed spheres with T1 values larger than 400 msec, because the implemented
B1+ map has low accuracy for lower T1 values [4]. The repeatability was
characterized as the coefficient of variation (CV) over 40 days. The
reproducibility between long and short MRF was evaluated by relative deviations
displayed as Bland-Altman plots. In addition, FISP-MRF was scanned in two human
volunteers (A and B). T1 and T2 values were obtained in 16 ROIs located in
different brain regions (see Figure 4), and repeatability and reproducibility
were evaluated.
Results
ISMRM/NIST system phantom: Figure 1 displays the T1 and T2 values and the temperature change among 40 days. There was no correlation between T1 or T2 value and the temperature. Figure 2 shows the 40-day CV; CVs were less than 0.7% for T1 values and less than 2.9% for T2 values. The 40-day variation between long and short MRF measurements of T1 and T2 is displayed in Figure 3. The mean bias for T1 was 0.4%, and the 95% limits of agreement ranged from -1.9% to 2.7%. The mean bias for T2 was 2.5%, and the 95% limits of agreement ranged from -6.0% to 1.0%. One data point with the shortest T2 value was outside of the limits of agreement.
Human Volunteers: Figure 4 shows the 5-day CV; CVs were less than 2.3% for T1 values and less than 5.1% for T2 values. The 5-day variation between long and short MRF measurements of T1 and T2 is displayed in Figure 5. The mean bias for T1 was -1.2%, and the 95% limits of agreement ranged from -4.6% to 2.3%. The mean bias for T2 was 4.9%, and the 95% limits of agreement ranged from -2.5% to 12.3%.
Discussion
Because the longer MRF version has twice the number
of acquisitions, there must be larger amount of information in the signal
evolution as compared to the shorter MRF version. Nevertheless, T1 and T2
values in the ISMRM/NIST phantom and healthy volunteers measured with both long
and short MRFs showed high repeatability. In the phantom and in the normal
brain tissue, there is comparable reproducibility between long and short MRF.
Short MRF showed the potential to be an equivalent substitute for long MRF in
terms of precision and accuracy. However, short MRF might underestimate T2
values around 20 msec and shorter. Conclusion
Acknowledgements
References
- Russek SE, Boss M, Jackson EF, Jennings DL, Evelhoch JL, Gunter JL, Sorensen AG. Characterization of NIST/ISMRM MRI system phantom. In Proceedings of the 20th Annual Meeting of ISMRM, Melbourne, Victoria, Austraia, 2012. Abstract 2456
- Jiang Y, Ma D, Seiberlich N, Gulani V, Griswold MA. MR fingerprinting using fast imaging with steady state precession (FISP) with spiral readout. Magn. Reson. Med. 2015;74:spcone-spcone. doi: 10.1002/mrm.26048.
- Pfeuffer J, Kechagias A, Meyer CH, Körzdörfer G, Nittka M. Mitigation of Spiral Undersampling Artifacts in Magnetic Resonance Fingerprinting (MRF) by Adapted Interleave Reordering. In Proceedings of the 25th Annual Meeting of ISMRM, 2017. Abstract 0133
- Chung S, Kim D, Breton E, Axel L. Rapid B1+ mapping using a preconditioning RF pulse with turboFLASH readout. Magn. Reson. Med. 2010;64:439–446. doi: 10.1002/mrm.22423.
Figures
