Yun Jiang1, Jesse I. Hamilton1, Katherine L. Wright2, Dan Ma2, Nicole Seiberlich1, Vikas Gulani1,2, and Mark A. Griswold1,2
1Department of Biomedical Engineering, Case Western Reserve University, Clevleand, OH, United States, 2Department of Radiology, Case Western Reserve University, Clevleand, OH, United States
Synopsis
The purpose of this study is to develop a method for simultaneous quantification of T1, T2, and diffusion within the MR Fingerprinting (MRF) framework. Multiple diffusion-weighted driven-equilibrium modules are inserted into the MRF-FISP acquisition in this study.
The results from the prostate show the
promising ability of this combination of MRF-FISP and diffusion preparation
modules to quantify relaxation parameters along with diffusion in one scan.Purpose
Quantitative relaxometry has shown promise for characterizing pathology
1 and monitoring treatment effects
2. Similarly, quantitative diffusion is also an important biomarker for disease diagnoses
3. These tissue properties (T
1,
T2 and apparent diffusion coefficient) are usually acquired in separate scans. Motion and other effects limit our ability to analyze these multidimensional quantitative data because of misregistration between different property maps. In particular, diffusion quantification is commonly acquired by diffusion–weighed spin echo EPI sequences which are prone to image distortions due to the long EPI readout. Non-rigid image registration would be necessary to obtain tissue properties coaligned within the same region-of-interest among different scans, and these distortions further complicate the analysis. The purpose of this study is to develop a method for simultaneous quantification of T
1, T
2, and diffusion within the MR Fingerprinting (MRF)
4 framework. MRF has been shown to be efficient in generating T
1 and T
2 maps by matching transient-state signals to a pre-calculated dictionary. The previous proposal
5 that relies on steady-state imaging sequences is sensitive to macroscopic motion during the readout. Inspired by the previous studies that added multiple preparation modules in MRF sequence
6,7, multiple diffusion-weighted driven-equilibrium modules
8 are inserted into the MRF-FISP acquisition
9 in this study in order to overcome these potential sources of error due to macroscopic motion. We show that this provides a signal evolution that is sensitive to diffusion as well as T
1 and T
2 relaxation in a single rapid imaging acquisition.
Methods
Figure 1 shows a diagram of the flip angles of the proposed MRF method. In the first 1045 images, the original MRF-FISP acquisition is executed with the sinusoidal varying flip angles ranging from 5° to 45° and varying repetition times from 6.9 ms to 10 ms. In order to increase the diffusion sensitivity, diffusion preparation modules with b-values of 50, 500 and 1000 s/mm
2 were inserted every 100 images. Images after the diffusion module were acquired with the sinusoidal varying flip angles ranging from 1° to 15° and a fixed repetition time of 6.9 ms. In this work, one arm of a uniformly sampled spiral trajectory that requires 48 interleaves to fully sample a 256 x 256 matrix size and 400 mm FOV was used to acquire each image. The spiral trajectory rotates 7.5° from one image to the next. A dictionary of the possible signal evolutions with a range of T
1 (20 ms - 3000 ms), T
2 (10 ms - 500 ms) and ADC (0 - 3×10
-3 mm
2/s) was simulated using the Bloch simulations. We employed a template-matching algorithm to match the obtained signal evolution to the closest dictionary entry and thus simultaneously extract the corresponding T
1, T
2, ADC and proton density values. For the proof of concept, patients with elevated Prostate-Specific Antigen (PSA) were scanned on a Siemens Magnetom Skyra 3T (Siemens AG Medical Solutions, Erlangen, Germany) with an 18-channel body array and a 12-channel spinal array in an IRB-approved, HIPAA-compliant study. T
2-weighed images were acquired by TSE with TR of 7200 ms, TE of 96 ms, the spatial resolution of 0.6 × 0.6 mm
2, and the slice thickness of 3 mm for the anatomical reference. A.spin-echo, diffusion-weighted EPI images (b=50, 500 and 1000 s/mm
2) were acquired with 1.2 × 1.2 mm
2, and slice thickness of 3 mm for the reference of ADC values. The proposed MRF method was averaged 6 times, yielding a total acquisition time of 2 minutes and 30 seconds per slice.
Results
Figure 2 shows the clinical T2-weighted image and ADC map from spin echo diffusion-weighted EPI. The white arrow points to cancer in the left peripheral zone. Figure 3 shows the T
1, T
2 and ADC maps from the proposed MRF method. The mean T
1 and T
2 values of the Normal peripheral zone (NPZ) are 2070 ms and 125 ms, which are in the range of values reported in the previous literature
10. The mean T1 and T2 values of the lesion in the left peripheral zone
is 1150 ms and 50 ms. The mean ADC value from the proposed MRF method is 1.86 ± 0.44 mm
2/s for
NPZ, and is 0.73±0.17 mm
2/s for the lesion. The ADC values from the spin echo diffusion EPI are 1.9±0.3 mm2/s and 0.51±0.38 mm2/s for NPZ and lesion, respectively.
Discussion
This
work demonstrates an MRF method to quantify relaxation parameters together with
diffusion within the MRF framework that shows reduced sensitivity to
macroscopic motion and image distortions than previous methods. The results
from the prostate show the promising ability of this combination of MRF-FISP
and diffusion preparation modules to quantify relaxation parameters along with
diffusion in one scan.
Acknowledgements
The
authors would like to acknowledge funding from Siemens Medical Solutions and
NIH grants 1R01EB017219, 1R01EB016728.References
1. Roebuck JR. et al. MRI 2009.
2. Weidensteiner C. et al. BMC Cancer 2014.
3. Padhani AR. et al. Neoplasia. 2009 Feb;11(2):102-25.
4. Ma D. et. al. Nature (2013) 495, 187–192
5. Jiang Y. et al. ISMRM 2014 pg#28
6. Hamilton J. et al. ISMRM 2015 pg#26
7 Anderson C. et al. ISMRM 2015 pg#3387
8 Thomas DL. 1998 Jun;39(6):950-60
9. Jiang Y. et. al. Magn. Reson. Med. 2014. doi: 10.1002/mrm.25559
10. Weis J. et. al. Magn Reson Med Sci, Vol. 12, No. 4, pp. 289–296, 2013