Buy One, Get Two for Free: Simultaneous Knee T2 Mapping and Morphological Analysis On Synthetic Images Using GRAPPATINI
Patrick Omoumi1, Tom Hilbert1,2,3, Marion Roux1, Jean-Baptiste Ledoux1, Ruud B Van Heeswijk1, Reto Meuli1, and Tobias Kober1,2,3

1Radiology, Lausanne University Hospital, Lausanne, Switzerland, 2Advanced Clinical Imaging Technology (HC CMEA SUI DI BM PI), Siemens Healthcare AG, Lausanne, Switzerland, 3LTS5, Ecole Polytechnique Fédérale de Lausanne, Lausanne, Switzerland

Synopsis

A fast quantitative T2 mapping technique that additionally provides synthetic images for morphological assessment was validated by two experienced radiologists regarding (1) the T2 values through a phantom experiment and (2) the image quality through a quantitative and qualitative assessment for the knee joint in five healthy volunteers.

Introduction

Today, intermediate-weighted (IW) and T2-weighted (T2w) Turbo Spin Echo (TSE) sequences with fat suppression are most commonly used to clinically assess morphological abnormalities of joint structures1. In addition, quantitative analysis of the musculoskeletal system using MR relaxometry techniques such as T2 mapping has gained interest in recent years. This is in particular the case for imaging of osteoarthritis, since T2 values can be used as a non-invasive biomarker of early degenerative disease of cartilage and meniscus2-4. The acquisition of both morphological and quantitative sequences is however time consuming. We suggest using a fast quantitative T2 mapping technique that allows the generation of synthetic images with different TEs, allowing a significant decrease in acquisition time while preserving image quality in comparison to the consecutive acquisition of morphological and quantitative sequences.

Methods

A multi-echo spin-echo (MESE) sequence was modified in order to acquire undersampled k-spaces. Generalized autocalibrating partially parallel acquisition (GRAPPA)5 and Model-based Accelerated Relaxometry by Iterative Non-linear Inversion(MARTINI)6 can be subsequently applied to estimate the transverse relaxation T2 and the equilibrium magnetization M0, a method termed GRAPPATINI7. Synthetic TSE images with any T2-weighting can then be generated using the M0/T2 maps in the forward signal model.

Phantom experiments were performed to validate the T2 estimation. To that end, the prototype GRAPPATINI and product MESE sequences were used to estimate the T2 values within tubes with different concentrations of Gadolinium and Agar using the same acquisition parameters as in the in-vivo experiments. A single-slice single-spin-echo product sequence was used to achieve reference T2 values using a standard log-linear fit onto various fully sampled acquisitions with different TEs=12,24,36,60,100ms.

Subsequently, the prototype GRAPPATINI sequence was used to estimate T2 and M0 maps of the knee joint at 3T (MAGNETOM Skyra, Siemens Healthcare, Germany) using a 15-channel knee coil in five healthy volunteers (3 males, age 30.2±3.3 years). Additional synthetic contrasts with TE=34ms and TE=80ms were generated on the scanner. For comparison, standard IW (TE=34ms) and T2w (TE=80ms) morphological TSE images were acquired. The detailed acquisition parameters are listed in Table 1.

The synthetic morphological images were validated quantitatively and qualitatively in comparison to the conventional TSE images. ROIs of at least 15mm2 were placed on fluid, muscle, meniscus and cartilage and copy-pasted between comparative images. SNR and CNR (cartilage/fluid and meniscus/fluid) were calculated. Qualitative analysis was performed by two radiologists in consensus blinded to the employed sequence by comparing the synthetic images and the corresponding TSE side-by-side in random order. A five-grade scale was used for the comparison (-2: first image significantly worse than second, -1: moderately worse, 0: no difference, +1: moderately better, +2: significantly better). Each of the following anatomical structures was assessed: cartilage, menisci, cruciate ligaments, bone marrow, muscle, joint fluid, quadricipital and patellar tendons. Furthermore, image contrast, noise, artifacts, and global diagnostic value were also compared.

Results and Discussion

Fig. 1 shows the T2 values of the phantom experiment. T2 values found by GRAPPATINI are slightly overestimated in comparison to the reference method, which is most likely due to stimulated echoes, a typical problem for T2 mapping using MESE sequences. The fully-sampled MESE sequence experiences a stronger overestimation which can be explained by the much shorter TR (1.6s versus 4.88s) causing an even stronger stimulated-echo effect due to increased T1 influences in the signal decay8.

The quantitative analysis showed similar SNR and no statistically significant difference between the synthetic and conventional sequences (average SNR=9.9 for both sequences, p=0.99). CNR values were not statistically different between the two sequences (cartilage/fluid: 6.2 vs. 6.6, p=0.62; meniscus/fluid: 11.3 vs. 11.6, p=0.81).

The qualitative analysis showed no difference in global image quality (cf. Fig 2) or of any of the anatomical structures that were evaluated (average score of 0, 95%CI=[0; 0.4]). Artefact scores were slightly higher for the synthetic sequences (average of -0.1, 95%CI=[-0.002;-0.6]), while visual noise and contrast were slightly better for the synthetic sequences (average score of 0.1, 95%CI=[0.002;0.6]).

Conclusions

The GRAPPATINI sequence provides accurate T2 values, as well as synthetic sequences that are quantitatively and qualitatively similar to conventional TSE sequences. Using this technique, T2 maps, IW and T2w sequences can all be obtained in 6.22min compared to 12.13min, corresponding to the sum of the acquisition times obtained with the standard technique.

Acknowledgements

No acknowledgement found.

References

1. Huang M, and Schweitzer ME. The role of radiology in the evolution of the understanding of articular disease. Radiology 2014;273:S1-22. 2. Kijowski R, Blankenbaker DG, Munoz Del Rio A, Baer GS, and Graf BK. Evaluation of the articular cartilage of the knee joint: value of adding a T2 mapping sequence to a routine MR imaging protocol. Radiology 2013;267:503-13. 3. Zarins ZA, Bolbos RI, Pialat JB, Link TM, Li X, Souza RB, and Majumdar S. Cartilage and meniscus assessment using T1rho and T2 measurements in healthy subjects and patients with osteoarthritis. Osteoarthritis Cartilage 2010;18:1408-16. 4. Rauscher I, Stahl R, Cheng J, Li X, Huber MB, Luke A, et al. Meniscal measurements of T1rho and T2 at MR imaging in healthy subjects and patients with osteoarthritis. Radiology 2008;249:591-600. 5. Griswold, Mark A., et al. Generalized autocalibrating partially parallel acquisitions (GRAPPA). Magnetic resonance in medicine 47.6 (2002): 1202-1210. 6. Sumpf, Tilman J., et al. Model-based nonlinear inverse reconstruction for T2 mapping using highly undersampled spin-echo MRI. Journal of Magnetic Resonance Imaging 34.2 (2011): 420-428. 7. Hilbert, Tom, et al. MARTINI and GRAPPA-When Speed is Taste. Proc. Intl. Soc. Mag. Reson. Med.. 22.4077 (2014). 8. Maier, Cynthia F., Steve G. Tan, Hari Hariharan, and Hollis G. Potter. T2 quantitation of articular cartilage at 1.5 T. Journal of magnetic resonance imaging 17, no. 3 (2003): 358-364.

Figures

Table. 1: Acquisition parameters of the GRAPPATINI sequence in comparison to the acquisition parameters of subsequently acquiring the different contrasts.

Fig. 1: Estimated T2 values using different methods within tubes containing different concentration of Gadolinium (Gd) and Agar gel showed in a table and in an agreement plot.

Fig. 2: Synthetic image (TE=34ms) derived from GRAPPATINI sequence with fat suppression (a) compared to conventional fat-suppressed TSE IW sequence (TE=34ms)(b). No significant difference is seen in terms of subjective image quality, or for the depiction of anatomical structures.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
4488