Synopsis

Mapping of quantitative MRI relaxation values is promising for improving the assessment of MSK disease. Magnetic Resonance Fingerprinting (MRF) is a new method that enables fast quantitative MRI by exploiting the transient signals caused by the variation of pseudorandom sequence parameters.

This proof-of-concept work demonstrates the utility of MR Fingerprinting in the knee. Seven participants, four of which had Kellgren-Lawrence (KL) grade 2 or 3, were imaged eighty minutes after gadolinium injection with MRF on a 3.0T MRI. The mean T1 relaxation times were shorter in cartilage by 5-20% in KL=2,3 subjects when compared to normal subjects.

Introduction

Mapping of quantitative MRI relaxation values is promising for improving the assessment of MSK disease (1). Magnetic Resonance Fingerprinting (MRF) is a new method that enables fast quantitative MRI by exploiting the transient signals caused by the variation of pseudorandom sequence parameters (2). These transient signals are then matched to a simulated dictionary of T1 and T2 values to create quantitative maps.

Cartilage macromolecules, such as glycosaminoglycans (GAGs), have a negative charge, which allows for the measurement of healthy and diseased tissue with a gadolinium chelate. For delayed gadolinium enhancement MRI of cartilage (dGEMRIC) (3), negatively charged gadolinium chelates accumulate in diseased cartilage where macromolecule concentration is lower than in healthy tissue, causing T1 and T2 shortening.

This work investigates MRF for T1 and T2 dGEMRIC mapping.

Methods

Seven participants were imaged on a 3.0 T MRI system (MR750 GE Healthcare, Waukesha, WI, USA) using an 8-channel transmit/receive knee coil. All imaging occurred with local ethical approval. Three participants had no history of OA, two participants had Kellgren-Lawrence (KL) grade 2, and two had KL grade 3. The subjects with KL 2-3 demonstrated predominant disease in the medial tibiofemoral compartment.

The MR protocol consisted of standard clinical sequences and multi-flip angle 3D spoiled gradient-echo (SPGR) T1-mapping, followed by gadolinium injection (Dotarem 0.4 mL/kg), ten minutes of exercise on a stationary cycle, and 80 minutes of rest. Following rest, subjects were imaged with a 3:09min 3D SPGR with 3 flip angles and a 4:40 minute 2D steady-state-free-precession (SSFP) MRF sequence (2, 4). SPGR parameters were: flip angles=6°,2°,14°, voxel size=0.5x0.5x3.0 mm3, matrix=320x256x40, NEX=0.5, field-of-view=160x128x120mm3, TR/TE=3.9/2.1ms, averages=2. The MRF acquisition acquired 979 frames with golden-angle spiral interleaves with field-of-view=225x225x63mm2, matrix = 192x192x18, voxel size=1.2x1.2x3.0mm3, slices=21, slice thickness=3.0mm, spacing 1.0mm, sampling bandwidth=±250kHz, slice dephasing=8π, TE=1.8ms, with repetition time and flip angle lists matching the values in Jiang (4). The maximum gradient strength per spiral was 30mT/m and the maximum slew rate was 105mT/m/s.

MRF maps were obtained by inner-product pattern matching of the dictionary with the reconstructed time frames. The MRF dictionary was computed for T1 and T2 using the extended phase graphs formalism (5) and included the slice profile (6). Synthetic images were created from the MRF generated maps using the spin-echo equation with TE=75ms for T2-weighting, and the gradient-echo equation with TR=220ms for T1-weighting.

Regions-of-interest (ROIs) were collected over the whole lateral femoral and tibial cartilage of all subjects. The mean and standard deviations (SDs) were calculated across subjects using the mean relaxation values of each individual.

Results

T1 and T2 maps generated from MRF are shown in Figures 1&2. Plots showing the means and standard deviations are in Figure 3. Synthetic images are shown in Figure 4.

The T1 means±SDs in the whole tibia (Normal/KL2/KL3) were 435.4±70.9/348.5±0.9/344.6±50.6ms and whole femur were 385.8±2.9/347.7±13.9/364.3±3.0ms. The T2 means in the whole tibia were 61.9±12.2/72.3±6.1/51.3±2.4ms and whole femur were 70.7±15.2/67.7±2.5/60.4±0.7ms.

Discussion

The mean T1 relaxation times were shorter by 5-20% in the KL=2 or KL=3 subjects when compared to normal subjects. The T2 means were not lower in KL>=2 subjects when compared to normal subjects, however, they were lower by >10% in KL=3 subjects when compared to KL=2 subjects.

Future MRF implementations could improve acquisition time or resolution by reducing the number of acquired frames, increasing spiral undersampling, applying 3D trajectories, and using compressed sensing. The traditional T1 mapping provided higher resolution than MRF. However, MRF has the advantage that it has the ability to simultaneously model multiple parameters, such as B1, which may be reducing the T2 values in the image peripheries.

The mean T2 relaxation times in the tibia were higher in KL=2 subjects and lower in KL=3 subjects, when compared to normal subjects, which does not follow the theory that contrast accumulates with increasing KL grade. This could be due to the relatively low numbers of this study (N=7), or due to partial volume effects from the large MRF voxel dimensions.

This study also measured whole lateral compartments, which will not be specific to an individual subject’s disease. This study demonstrated the feasibility of using MRF in osteoarthritic subjects, which is attractive due the ability of MRF to obtain multiple measurements within a single acquisition for analysis and image synthesis.

Conclusion

Acknowledgements

This was a University of Cambridge sponsored study with funding provided by GlaxoSmithKline.

The authors also acknowledge research support from the National Institute of Health Research Cambridge Biomedical Research Centre.

A travel award was funded by the Royal Society for collaboration between Pisa and Cambridge.

References

1. Mosher TJ, Dardzinski BJ, Smith MB. Human Articular Cartilage: Influence of Aging and Early Symptomatic Degeneration on the Spatial Variation of T2—Preliminary Findings at 3 T. Radiology. 2000;214(1):259-66.

2. Ma D, Gulani V, Seiberlich N, Liu K, Sunshine JL, Duerk JL, et al. Magnetic resonance fingerprinting. Nature. 2013;495(7440):187-92.

3. Bashir A, Gray ML, Burstein D. Gd-DTPA2− as a measure of cartilage degradation. Magnetic Resonance in Medicine. 1996;36(5):665-73.

4. Jiang Y, Ma D, Seiberlich N, Gulani V, Griswold MA. MR fingerprinting using fast imaging with steady state precession (FISP) with spiral readout. Magnetic Resonance in Medicine. 2015;74(6):1621-31.

5. Weigel M. Extended phase graphs: dephasing, RF pulses, and echoes‐pure and simple. Journal of Magnetic Resonance Imaging. 2015;41(2):266-95.

6. Buonincontri G, Sawiak SJ. MR fingerprinting with simultaneous B1 estimation. Magnetic Resonance in Medicine. 2016;76(4):1127-35.