Correlation Time Diffusion Brain Mapping at 1.5T vs. 3.0T
Hernan Jara1, Arnaud Guidon2, Lloyd Estkowski3, Jorge A Soto1, and Osamu Sakai1

1Radiology, Boston University, Boston, MA, United States, 2Global MR Applications and Workflow, GE Healthcare, Boston, MA, United States, 3Global MR Applications and Workflow, GE Healthcare, Menlo Park, CA, United States

Synopsis

Purpose: To test the hypothesis that correlation time diffusion coefficient (DCT) MRI is not be dependent of the main magnetic field strength B0. Methods: A heathy volunteer was scanned (head) at 1.5T and 3.0T with the same diffusion mapping capable scans: quadra-FSE for DCT mapping and DW-SE-sshEPI for DPFG mapping. Results: Despite the B0 dependent shifts of the relaxation time (T1, T2) histograms, the DCT, DPFG, and PD histograms do not exhibited significant B0 shifts. Conclusion: The correlation time diffusion coefficient is theoretically and experimentally a genuine physical property inherent to biological tissue and independent of experimental variables including B0.

Introduction

Quantitative MRI parameters include measures of 1) tissue hydration –the proton densities (PD) of the mobile and the semisolid water pools--, 2) tissue micro- and macro-kinetics (diffusion coefficient, perfusion parameters, and flow velocity), and 3) measures of spin interactions with the microscopic magnetic milieu via the relaxation times (T1, T2, T2*). Of these qMRI parameters, only the relaxation times depend on the main magnetic field strength B0 and the others are purely determined by the physical and physiological state of tissue. Accordingly, PD as well as the diffusion coefficient, as measured by the Stejskal-Tanner pulsed-field-gradient (PFG) experiment (1), are theoretically and experimentally independent of B0. Correlation time diffusion is an alternative qMRI technique for mapping the diffusion coefficient (2-4) --termed DCT to distinguish it from DPFG-- that uses relaxometric (T1 and T2) and proton density (PD) information (5-10). Hence, although DCT is theoretically B0 independent, the question arises as to whether the field independence holds in practice. To help elucidate this question, a volunteer was scanned at 1.5T and 3.0T with the same multispectral qMRI technique using the quadra-FSE pulse sequence and with the standard PFG diffusion technique.

Methods and Materials

The volunteer (male 35yo) was scanned at 3.0T (Discovery MR750, GE Healthcare, Waukesha, WI) and a week later at 1.5T (Optima MR450w, GE Healthcare, Waukesha, WI). The imaging protocol included: a) standard PFG diffusion-weighted scan using the DW single-shot EPI sequence, and b) a newly developed pulse sequence termed herein quadra fast spin-echo (quadra-FSE; see Figure 1), which is multispectral in PD, T1, and T2 and therefore can be used for DCT mapping. The DW-SE-sshEPI sequence was implemented with TE=63.9ms and TR=8,000ms) and diffusion encoding along the three physical axes (b=0 and 1,000s/mm2). The quadra-FSE pulse sequence is a concatenation of two dual-echo fast spin-echo (DE-FSE) that are run consecutively with equal parameters including pre-scan values, but with different TR values and consequently different levels of T1 saturation. The key pulse sequence parameters are the two TRs (5,320 & 600ms) and the two effective echo times (TE1eff=22.1 & TE2eff=99.6ms with ETL=10). Four directly acquired images per slice are generated with varying levels of T1- and T2-weighting; specifically Sat1_E1, Sat1_E2, Sat2_E1, and Sat2_E2 (see Figure 1) and these images can be qMRI processed to generate coregistered maps of PD, T1, and T2. In turn, these maps were used to map DCT using equations derived from the microscopic theory of relaxation. The directly acquired images of both scans were transferred to a personal computer equipped with Mathcad (version 2001i, PTC, Needham, MA) and qMRI processed with model conforming algorithms to produce maps of PD, T1, T2, DCT, and DPFG. These maps were segmented with a dual-clustering algorithm leading to datasets representing the totality of the intracranial tissues (white matter (WM), gray matter (GM), meninges, and cerebrospinal fluid (CSF)). The resulting intracranial matter (ICM) segments were further processed with a pixel counting algorithm to generate histograms representing the individual distributions of each qMRI parameter.

Results

ICM T1 histograms representing all-intracranial tissues of the same volunteer scanned at 1.5T (blue line) and 3.0T (red line) with the quadra-FSE pulse sequence are shown in Figure 2. These are fundamentally tri-modal histograms with well-defined spectral features representing WM, GM, and CSF. Note that as is well known, the WM and GM T1values at 3.0T are longer than at 1.5T by 15%-20%. Analogously, T2 histograms of the all-intracranial tissues of the same volunteer scanned at 1.5T (blue line) and 3.0T (red line) with the quadra-FSE pulse sequence are shown in Figure 3. In this case the spectral shift is in the opposite direction; the T2 values at 3.0T are shorter than at 1.5T by 15%-20%. Despite the B0 dependent shifts of the relaxation time histograms, the PD, DCT, and DPFG histograms do not exhibited significant B0 shifts (see Figures 4 & 5).

Conclusion

These experiments performed on the same human subject confirm the hypothesis that the physico-physiological qMRI parameters (PD, DCT, and DPFG) are independent of the main magnetic field strength, within experimental uncertainty. Most importantly, this study further confirms the physical interpretation of DCT as a purely physical and B0-independent diffusion measure, despite the fact that it is calculated using the relaxation times, which are B0 dependent.

Acknowledgements

This work was supported in part by a research grant from GE Healthcare.

References

1. Stejskal E, Tanner J. Spin diffusion measurements: spin echoes in the presence of a time-dependent field gradient. Journal of Chemical Physics, Vol 42 1965:288-292.

2. Price W. Pulsed-field gradient nuclear magnetic resonance as a tool for studying translational diffusion: Part 1. Basic theory. Concepts in Magnetic Resonance Part A 1997;9(5):299-336.

3. Jara H. High spatial resolution diffusion-MRI of the human brain with the mixed-TSE pulse sequence: a non-pulsed field gradient technique. Proceeding of the RSNA Chicago, IL 2005.

4. Jara H. Correlation time diffusion coefficient brain mapping: combined effects of magnetization transfer and water micro-kinetics on T1 relaxation. Proceedings ISMRM (Toronto, Canada) 2008. 5. Bloembergen N, Purcell E, Pound R. Relaxation effects in nuclear magnetic resonance absorption. Physical Review, vol 73 1948(7):679-712.

6. Torrey H. Nuclear spin relaxation by translational diffusion. Physical Review, vol 92 1953(4):962-969.

7. Solomon I. Relaxation processes in a system of two spins. Physical Review 1955;99(2):559-565.

8. Resing H, Torrey H. Nuclear spin relaxation by translational diffusion. III. Spin-spin relaxation. Physical Review, vol 131 1963(3):1102-1104.

9. Bottomley P, Foster T, Argersinger R, Pfeifer L. A review of normal tissue hydrogen NMR relaxation times and relaxation mechanisms from 1–100 MHz: dependence on tissue type, NMR frequency, temperature, species, excision, and age. Medical Physics 1984;11:425.

10. Gore JC, Kennan RP. Physical and physiological basis of magnetic relaxation. Magn Reson Imaging 1999;1:33-42.

Figures

Figure 1. The herein termed quadra-FSE pulse sequence in its simplest form is the concatenation of two DE-FSE acquisitions with different TR values that are run with the same pre-scan parameters. This pulse sequence uses two different levels of T1 saturation and two levels of T2 weighting thus enables the generation of T1, T2, and PD maps.

Figure 2. T1 histograms of the all-intracranial tissues of the same volunteer scanned at 1.5T (blue line) and 3.0T (red line) with the quadra-FSE pulse sequence. Note that as is well known, the T1 values at 3.0T are longer than at 1.5T by 15%-20%.

Figure 3. T2 histograms of the all-intracranial tissues of the same volunteer scanned at 1.5T (blue line) and 3.0T (red line) with the quadra-FSE pulse sequence. Note that as is well known, the T2 values at 3.0T are shorter than at 1.5T by 15%-20%.

Figure 4. PD histograms of the all-intracranial tissues of the same volunteer scanned at 1.5T (blue line) and 3.0T (red line) with the quadra-FSE pulse sequence.

Figure 5. DCT and DPFG histograms of the all-intracranial tissues of the same volunteer scanned at 1.5T (blue lines) and 3.0T (red lines) with the quadra-FSE and with DW-single-shot SE-EPI pulse sequences. Notably, the four histograms peak at approximately the same value (~0.8 10-3mm2/s) and the two DCT histograms are considerably narrower than their DPFG counterparts are.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
3048