Synopsis

Stimulated-echo diffusion-imaging method was employed to reduce the effect of macroscopic field inhomogeneity and large vessel overestimation on the quantification of blood volume fraction and mean vessel radius. Stimulated-echo diffusion-imaging method were compared with conventional spin-echo and gradient-echo methods by Monte Carlo (MC) proton diffusion simulations and in vivo rat experiments on a 7 T system. The results of this study showed that stimulated-echo based MR relaxation-rates, ∆R_{STE,long TD} (ΔR₂^* like) and ∆R_{STE,short TD} (ΔR₂ like), provide the robust means of assessing cerebral microvasculature where the macroscopic field inhomogeneity is severe and signal contamination from adjacent large vessel is significant without necessity of co-registering gradient- and spin-echo images.

Introduction

Measurements of changes in magnetic resonance (MR) relaxation-rates, ΔR₂ and ΔR₂^*, induced by the injection of an intravascular contrast media enable the quantification of cerebral microvasculature such as blood volume fraction and mean vessel radius.^1,2 However, the macroscopic field inhomogeneity from air-tissue interface and large blood vessel results in measurement error of ΔR₂^* due to gradient echo acquisition.^1-3 Also, co-registering maps from separate gradient- and spin-echo acquisitions typically suffer from different point spread functions.

In this study, stimulated-echo diffusion-imaging method was employed to reduce the effect of macroscopic field inhomogeneity and large vessel overestimation on the quantification of blood volume fraction and mean vessel radius. Stimulated-echo diffusion-imaging method were compared with conventional spin-echo and gradient-echo methods by Monte Carlo (MC) proton diffusion simulations and in vivo rat experiment on a 7 T system.

Methods

Finite perturber method was employed to calculate the B₀ shift of randomly distributed cylinder models (0.4 × 0.4 × 0.4 mm³) with various radii (2 to 30 μm) and volume fractions (2, 4, 6, and 8 %).⁴ Δχ for the simulation was 3.6 ppm (SI unit, dose of 200 μmol iron/kg when using USPIO).¹ Stimulated-echo, spin-echo, and gradient-echo acquisitions were simulated by MC approach. Echo times (TE) for stimulated-echo, spin-echo, and gradient-echo acquisitions were 10 ms, and diffusion times (TD) for stimulated-echo acquisition were 10 and 1200 ms. Changes of transverse relaxation-rates (ΔR₂, ΔR₂^*, and ΔR_STE for spin-echo, gradient-echo, and stimulated-echo respectively) were calculated by following equation:

ΔR₂, ΔR₂^*, and ΔR_STE = 1/TE ln(S_pre/S_post) (1)

where S_pre and S_post are the signal intensity before and after administration of contrast agent respectively. Vessel size index based on spin-echo/gradient-echo and stimulated-echo methods were calculated by ΔR₂^*/ΔR₂ and ΔR_{STE,TD=1200ms}/ΔR_STE,TD=10ms respectively.

For the in vivo experiment, stimulated-echo sequence with phase cycling scheme was implemented on the Bruker 7 T scanner.⁵ To measure cerebral blood volume fraction and vessel size index of normal Wister rat (208 g), ΔR₂^*, ΔR₂, ΔR_STE,TD=10ms, and ΔR_{STE,TD=1000ms} was measured by spin-echo, gradient-echo, and stimulated-echo acquisitions with following parameters: pulse repetition time = 2500 ms, TE = 10 ms, number of scans = 4, acquisition resolution = 0.234 × 0.234 × 1 mm³, contrast agent dose of 360 μmol iron/kg using intravascular superparamagnetic iron oxide nanoparticles (SPION).

Results

Representative volume rendered images of randomly distributed cylinder models are shown in Figure 1A. Randomness of the cylinder distribution decreases as the cylinder radius increases. Figure 1B illustrates computed changes of relaxation-rates. ΔR₂^* and ΔR_{STE,TD=1200ms} are linearly proportional to the blood volume fraction, and ΔR₂ and ΔR_STE,TD=10ms decrease as the cylinder radius increases. ΔR₂^*/ΔR₂ and ΔR_{STE,TD=1200ms}/ΔR_STE,TD=10ms are proportional to the cylinder radius, and blood volume effect was compensated. For the small cylinder radius less than 10 μm, ΔR_{STE,TD=1200ms}/ΔR_STE,TD=10ms shows smaller standard deviation than ΔR₂^*/ΔR₂.

Figure 2 shows in vivo measurement of cerebral ΔR₂^*, ΔR₂, ΔR_STE,TD=10ms, and ΔR_{STE,TD=1000ms}, and ΔR₂^*/ΔR₂ and ΔR_{STE,TD=1000ms}/ΔR_STE,TD=10ms are illustrated in Figure 3. In vivo results are comparable to the results of MC simulation. As pointed with white arrows, ΔR₂^* is distorted where the macroscopic field inhomogeneity is severe, and the blood volume is overestimated where the large vessel exists. Whereas, the results from stimulated-echo based vessel size index appears to be robust again them.

Discussion

Acknowledgements

References

1. Troprès, Irène, et al. "Imaging the microvessel caliber and density: principles and applications of microvascular MRI." Magnetic resonance in medicine 73.1 (2015): 325-341.

2. Emblem, Kyrre E., et al. "Vessel calibre—a potential MRI biomarker of tumour response in clinical trials." Nature reviews Clinical oncology 11.10 (2014): 566.

3. Kiselev, Valerij G., et al. "Vessel size imaging in humans." Magnetic resonance in medicine 53.3 (2005): 553-563.

4. Pathak, A.P., B.D. Ward, and K.M. Schmainda, A novel technique for modeling susceptibility-based contrast mechanisms for arbitrary microvascular geometries: the finite perturber method. Neuroimage, 2008. 40(3): p. 1130-1143.

5. Sprinkhuizen, S. M., Ackerman, J. L., & Song, Y. Q. (2014). Influence of bone marrow composition on measurements of trabecular microstructure using decay due to diffusion in the internal field MRI: Simulations and clinical studies. Magnetic resonance in medicine, 72(6), 1499-1508.