Dual-pathway sequences for MR thermometry: When and where to use them
Pelin Aksit Ciris1, Cheng-Chieh Cheng2, Chang-Sheng Mei3, Lawrence P. Panych2, and Bruno Madore2

1Department of Biomedical Engineering, Akdeniz University, Antalya, Turkey, 2Department of Radiology, Brigham and Women’s Hospital, Harvard Medical School, Boston, MA, United States, 3Department of Physics, Soochow University, Taipei, Taiwan

Synopsis

Dual-pathway sequences have been proposed to improve the temperature-to-noise-ratio (TNR) in MR thermometry. The present work establishes how much improvement these sequences may bring for various tissue types. Simulation results were validated against analytical equations, phantom and in vivo human results. PSIF-FISP thermometry allowed TNR improvements for kidney, pelvis, spleen or gray matter, and up to 2-3 fold reductions in TR with 20% TNR gains were achievable. Further TNR benefits are expected for heated tissues, due to heating-related changes in relaxation rates. In other tissue types such as liver, muscle or pancreas improvements were observed only for short TR settings.

PURPOSE

To find out how much of a boost in temperature-to-noise-ratio (TNR) can be achieved with dual-pathway sequences, as a function of tissue type.

INTRODUCTION

MR thermometry is a valuable tool for monitoring thermal therapies. The most common approach is based on the proton-resonant-frequency (PRF) shift of water (1-3). The phase maps required for PRF thermometry are typically obtained using gradient-recalled echo (GRE) sequences. Dual-pathway unbalanced steady-state sequences have been proposed instead to boost TNR (4). These sequences sample a ‘fast imaging with steady-state free precession’ (FISP) signal late in the TR period and an inverted-FISP (PSIF) signal early in TR (4). The acquisition is similar to that of a dual-echo steady state (DESS) sequence (5), except for the order of the PSIF and FISP echoes. The basic idea is that the temperature sensitivity of the PSIF signal depends on (TR-TE) rather than TE, making FISP and PSIF signals natural partners toward utilizing the entire TR period in a TNR-effective manner, with one signal pathway sampled early and the other one late in TR.

However, the signal intensity of the PSIF signal greatly varies according to parameter setting and relaxation parameters, making it unclear how much of a TNR boost might be obtained. Using a two-echo FISP sequence as a reference standard, whereby the two FISP echoes have the same TE and bandwidth as the PSIF and FISP of the tested dual-pathway sequence, TNR boosts were quantified here for a number of different tissue types and parameter settings.

METHODS AND RESULTS

All in vivo and phantom experiments performed here were aimed primarily at validating our simulation program. Whenever possible, validation was also performed against analytical expressions. Once validated, this software could then be used to test a wide array of tissue types and imaging parameters.

Validation against analytical solutions ($$$R_2' = 0$$$): With no reversible decay component, i.e., $$$R_2'$$$=0 and $$${T_2}^*$$$=$$$T_2$$$, the analytical value for PSIF and FISP signal can be found in Refs (6-8). Analytical and simulated results for the strength of both signal types, normalized by $$$M_0$$$, were obtained over a wide range of $$$T_1$$$, $$$T_2$$$, flip angle and TR settings. Analytical and simulated results showed near-perfect agreement, see Fig. 1.

Validation against in-vivo and phantom experiments ($$$R_2' \neq 0$$$): Following informed consent, abdominal imaging was performed on three subjects and brain imaging on four subjects, on a 3 T system using the PSIF-FISP sequence. Phantoms doped with varying concentrations of manganese sulfate (MnSO4) were scanned as well. Predictions for the relative PSIF / FISP signal, as obtained from our simulation program, were compared to their in vivo and phantom counterparts (Fig. 2). Close agreement can be seen relative to the identity line (Fig. 2a, b and c), and in Bland-Altman plots (Fig. 2d, e, f, g).

Validation of TNR values: Simulated signal levels for PSIF and FISP pathways were converted into relative TNR, as compared to the dual-FISP reference standard. Monte Carlo simulations, whereby various levels of noise were added to data, were used to validate this conversion step from signal levels to relative TNR. As shown in Fig. 3, near perfect agreement was obtained between analytical and Monte Carlo values.

TNR-based recommendations: The purpose of Fig. 1-3 was to validate our simulation program, while the purpose of Fig. 4-5 was to generate recommendations on when/where to use PSIF-FISP for thermometry. Results are shown in Fig. 4 for many different tissue types (9-14), over a wide range of TR and flip angle settings. Green indicates a TNR boost while red indicates a TNR penalty for the PSIF-FISP sequence. Contour plots were overlaid to show how well the reference standard (dual-FISP) sequence performed as a function of TR and flip angle (‘100’ for maximum performance, contour ‘90’ indicates 90% of maximum performance, etc.). Figure 5 plots the TNR boost achieved with a PSIF-FISP sequence as a function of $$$T_1$$$ and $$$T_2$$$ (for TR = 15 ms, flip angle = 60$$$^o$$$).

DISCUSSION AND CONCLUSION

As seen in Fig. 4, wherever the reference standard dual-FISP sequence performs best (innermost contour) the PSIF-FISP sequence has very little more to offer. However, whenever TR is shortened enough, for faster imaging and/or improved motion robustness, the PSIF-FISP sequence becomes advantageous (see green regions in Fig. 4 at short TR settings). The PSIF-FISP sequence is generally better suited for tissues with longer $$$T_1$$$ and $$$T_2$$$ values (e.g. kidney), and not so much otherwise (e.g. liver). It is worth noting that as tissues are heated, both $$$T_1$$$ and $$$T_2$$$ tend to increase, making PSIF-FISP sequences more desirable (Fig. 5).

Acknowledgements

Financial support from NIH grants R25CA089017, R01CA149342, R01EB010195, R21EB019500 and P41EB015898 is duly acknowledged.

References

1. De Poorter J, De Wagter C, De Deene Y, Thomsen C, Stahlberg F, Achten E. Noninvasive MRI thermometry with the proton resonance frequency (PRF) method: in vivo results in human muscle. Magn Reson Med 1995;33(1):74-81.

2. Ishihara Y, Calderon A, Watanabe H, Okamoto K, Suzuki Y, Kuroda K, Suzuki Y. A precise and fast temperature mapping using water proton chemical shift. Magn Reson Med 1995;34(6):814-823.

3. Peters RD, Hinks RS, Henkelman RM. Ex vivo tissue-type independence in proton-resonance frequency shift MR thermometry. Magn Reson Med 1998;40(3):454-459.

4. Madore B, Panych LP, Mei CS, Yuan J, Chu R. Multipathway sequences for MR thermometry. Magn Reson Med 2011;66(3):658-668.

5. Bruder H, Fischer H, Graumann R, Deimling M. A new steady-state imaging sequence for simultaneous acquisition of two MR images with clearly different contrasts. Magn Reson Med 1988;7(1):35-42

6. Hanicke W, Vogel HU. An analytical solution for the SSFP signal in MRI. Magn Reson Med 2003;49(4):771-775.

7. Zur Y, Stokar S, Bendel P. An analysis of fast imaging sequences with steady-state transverse magnetization refocusing. Magn Reson Med 1988;6(2):175-193.

8. Gyngell ML. The steady-state signals in short-repetition-time sequences. Journal of Magnetic Resonance (1969) 1989;81(3):474-483.

9. Wansapura JP, Holland SK, Dunn RS, Ball WS Jr. NMR relaxation times in the human brain at 3.0 tesla. J Magn Reson Imaging. 1999 Apr;9(4):531-8.

10. de Bazelaire CM, Duhamel GD, Rofsky NM, Alsop DC. MR imaging relaxation times of abdominal and pelvic tissues measured in vivo at 3.0 T: preliminary results. Radiology. 2004 Mar;230(3):652-9.

11. Gold GE, Han E, Stainsby J, Wright G, Brittain J, Beaulieu C. Musculoskeletal MRI at 3.0 T: relaxation times and image contrast. AJR Am J Roentgenol. 2004 Aug;183(2):343-51.

12. Stanisz GJ, Odrobina EE, Pun J, Escaravage M, Graham SJ, Bronskill MJ, Henkelman RM. T1, T2 relaxation and magnetization transfer in tissue at 3T. Magn Reson Med. 2005 Sep;54(3):507-12.

13. Rakow-Penner R, Daniel B, Yu H, Sawyer-Glover A, Glover GH. Relaxation times of breast tissue at 1.5T and 3T measured using IDEAL. J Magn Reson Imaging. 2006 Jan;23(1):87-91.

14. Wright PJ, Mougin OE, Totman JJ, Peters AM, Brookes MJ, Coxon R, Morris PE, Clemence M, Francis ST, Bowtell RW, Gowland PA. Water proton T1 measurements in brain tissue at 7, 3, and 1.5 T using IR-EPI, IR-TSE, and MPRAGE: results and optimization. MAGMA. 2008 Mar;21(1-2):121-30.

Figures

Fig.1: Dual-pathway PSIF and FISP signals as a function of $$$T_1$$$, $$$T_2$$$, Flip angle, and TR, with $$$R_2' = 0$$$. Simulations (*) and analytical solutions (-) show excellent agreement.

Fig.2: Dual-pathway simulations vs. experiments, with $$$R_2 \neq 0$$$, agree well. Results for each pathway relative to the identity line and ROIs on sample images are shown for (a) Phantoms (b) Human abdomen (c) Human brain. (d-g) Bland-Altmann plots are also provided for phantom experiment vs. simulations.

Fig.3: Relative TNRs from TNR equations vs. Monte Carlo Experiments, for all simulated tissues and acquisition parameters.

Fig.4: Relative TNR of PSIF-FISP vs. dual-FISP for various tissues and acquisition parameters. Green: higher TNR with PSIF-FISP. Red: higher TNR with dual-FISP. Contours: TNR with PSIF-FISP (at each TR and FA parameter set), relative to the maximum achievable TNR with dual-FISP across all tested TR and FA parameter sets (percentage).

Fig.5: Relative TNR of PSIF-FISP vs. dual-FISP, as a function of $$$T_1$$$ and $$$T_2$$$ (at sample parameter values of TR = 15ms, FA = 60).



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
3001