Purpose

A new ultrafast gradient-echo-based 3D imaging method is proposed using spatiotemporal encoding^{1, 2} (SPEN), which is dubbed RASE (Rapid Acquisition with Sequential Excitation). Two versions of RASE (I and II) are available: One is to enable a much shorter effective TE than EPI counter parts and the other provides constant TE across an object to be imaged. Here the second one (RASE-II) is introduced and demonstrated by lemon and in-vivo rat brain imaging on a 9.4-T (Bruker-BioSpec, 94/30 US/R) animal scanner. A potential application for fMRI was also tested by examining tSNR maps from time series.

Method

The sequence diagram of RASE-II is illustrated in Fig.1a. A frequency-swept chirp pulse is used for sequential spin excitation which produces a quadratic phase that localizes a signal in both time and space during data acquisition in the presence of gradient. In RASE-II, spin excitation and data acquisition are performed with the same duration and, between them, a rephasing gradient is applied to make all the spins experience constant TE (Fig.1b). Also note that the slab-selective direction is spatiotemporally encoded, which allows a flexibility for reducing the echo train length and the effective TE³. For demonstration, high-resolution lemon and in-vivo rat brain imaging was performed using a single-channel transceiver volume coil with 72-mm diameter and a surface coil with 35-mm diameter, respectively. Images were reconstructed offline with MATLAB (ver.8.2.0; R2013b) using the super-resolution algorithm⁴. For lemon imaging, 2D-GRE, GE-EPI, and RASE-II sequences were used for comparison. Scan parameters in common were: FOV = 65×65 mm², matrix = 128×128, number of slices = 48, slice thickness (THK) = 0.3125 mm. In GRE imaging as a reference, TR/TE = 310/3 ms, flip angle (FA) = 33°. In GE-EPI imaging, volume-TR/TE = 3,600/36.6 ms, FA = 90°. For RASE-II imaging, shot-TR/TE = 57/29.5 ms, FA (Ernst angle) = 14.3°, R-value (pulse_length × bandwidth) = 256. In-vivo rat brain imaging was performed using GE-EPI and RASE-II. To evaluate temporal signal fluctuations, time series were acquired in the resting state with 60 volumes and tSNR maps were calculated by dividing the temporal mean value of each voxel by the temporal standard deviation. Same scan parameters were: volume-TR/TE = 2,400/20 ms, FOV = 30×15 mm², matrix = 96×48, number of slices = 48, THK = 0.3125 mm, FA (Ernst angle) = 75°(GE-EPI), 13.4°(RASE-II). In RASE-II imaging, shot-TR = 50 ms, R-value = 64.

Results

Figure 2 shows lemon images from 2D-GRE (a), GE-EPI (b), RASE-II (c). RASE-II provided better image quality with less distortion than GE-EPI because of higher tolerance to field inhomogeneites due to larger gradient amplitudes as well as less T₂^* effects due to local rephasing mechanism in the SPEN direction. Figure 3 shows axial images of in-vivo rat brain (a, c), their tSNR maps (b, d), and the histogram of tSNR distribution in the brain area (e) obtained from GE-EPI and RASE-II. It is clearly shown that tSNR of RASE-II was overall higher than that of GE-EPI, as it was previously reported that 3D imaging provides better tSNR than 2D counterparts in case that thermal noises are dominant compared to physiological noises, e.g., in high-resolution 3D imaging^{5, 6}.

Discussion & Conclusion

Despite a gradient-echo-based acquisition scheme, RASE-II inherits most of appealing features of other spin-echo-based SPEN imaging methods such as no Nyquist ghosting in the SPEN direction and higher immunity to geometric distortions due to field inhomogeneities than EPI counterparts^{1, 2}. In addition, low flip angles make RASE-II less sensitive to specific absorption rate (SAR) and B₁-inhomogeneity. Moreover, since RASE-II is able to provide constant TE across all spins, it is barely affected by T₂^* signal modulation seen in GE-EPI. With regard to fMRI, the 3D characteristics of RASE-II with low flip angle reduces inflow effects as well as physiological noises⁷. As shown in Fig.3, high-resolution RASE-II imaging significantly improves tSNR over GE-EPI in the resting-state fMRI scheme. In conclusion, RASE-II will be promising in many GE-EPI applications including fMRI, especially offering more benefits at ultrahigh fields, e.g., 7 T and above.

Acknowledgements

This work was supported by IBS-R015-D1

References

[1] Camberlain R et al., Mag Reson Med 2007;58:794-799. [2] Ben-Eliezer N et al., Mag Reson Imaging. 2010;28:77-86. [3] Ryu JK et al., ISMRM. 2015;6897. [4] Chen Y et al., Mag Reson Med. 2013;69:1326-1336. [5] Poser B.A at al., Neuroimage. 2010;51:261-266. [6] van der Zwaag w at al., Mag Reson Med. 2012;67:344-352. [7] Gao J.H at al., Neuroimage. 2012;62:1035-1039.