Introduction:

For reasons of speed and clinical utility it is conceptually appealing to pursue MRI techniques that are sensitive to multiple contrast mechanisms and support simultaneous generation of two or more contrasts [1]. Generating two or even more contrasts simultaneously within one acquisition, offers substantial reduction of measurement time. This is of benefit in a clinical scenario, but also in preclinical studies where multiple parameters are assessed using imaging protocols with tight spatio-temporal resolution constraints. Moreover, simultaneous acquisition of multiple contrasts obviates the need for slice co-registration when combining multiple series of images with different contrast, which would be advantageous for applications dealing with bulk or physiological motion such as cardiac or ophthalmic MRI. Realizing this opportunity this abstract proposes a dual contrast RARE-EPI hybrid [2], that provides T₂ (RARE module) and T₂* (EPI module) contrast and facilitates simultaneous T₂ and T₂* mapping in a single radially (under)sampled scan, which will be referred to 2-in-1 RARE-EPI for reasons of brevity. The applicability of 2-in-1 RARE-EPI is carefully examined in phantom and in in vivo studies and benchmarked versus conventional T₂ and T₂* weighted/mapping techniques.

Methods:

In the RARE-EPI hybrid [3] the first echoes in the echotrain are acquired with a RARE module whereas the later echoes are acquired with an EPI module (Figure 1, left). The combined-acquisition-technique (CAT) factor λ (λ=0.5 in this work) determines the fraction of RARE echoes in the echotrain. In this work, RARE and EPI echoes are acquired in an interleaved manner with respect to the angular position in radially sampled k-space (Figure 1, right). For the view ordering of the RARE echoes we used the bit-reverse scheme [4] to reduce T₂-decay induced streaking artefacts, but without distributing echoes in one TR over the full angular range. For the EPI module sequential view ordering was applied to ensure short echo spacing. The method proposed by Block et al. [5] was used to correct for gradient delay artefacts. Images from radial k-space data were reconstructed using regridding with linear density compensation and a magnitude sum-of-squares channel combination. To reconstruct images including all EPI echoes a phase correction (PC) [6] was used, since magnetic field inhomogeneities lead to phase inconsistencies between EPI spokes. Reference T₂ and T₂* maps were acquired with conventional multi-spin-echo (MSE) and multi-gradient-echo (MGRE) techniques (see caption of Figure 2 for sequence parameters). T₂ and T₂* maps were calculated with a linear least-square-fit of the logarithmic signal magnitude. For T₂* mapping a spatial-adaptive-non-local-means filter was applied to the images prior to the fitting to improve the fit quality [7]. Images were acquired at 3.0T (Siemens Magnetom Verio, Erlangen, Germany) using a 32-channel receive-only head coil (Siemens, Erlangen, Germany). Two phantoms were used: A spherical homogeneous phantom (diameter=15cm), filled with Dimethylpolysiloxan-oil (T₂=106ms and T₁=452ms), and cylindrical agarose tube phantom (outer diameter=12.5cm, tube diameter=1cm). In vivo images of the brain were acquired from a healthy subject.

Results:

Figure 2 shows MR images of the two phantoms and the brain of a healthy subject acquired with 2-in-1 RARE-EPI. The images reconstructed from all echoes, from RARE echoes only (T₂ weighted) and from EPI echoes (T₂* weighted) exhibit different contrasts. The phase correction (PC) [6] reduces the signal voids which occur if all EPI echoes are combined for image reconstruction. Figure 3 demonstrates, that the T₂ map (RARE module, mean T₂=113±2ms) of the uniform Dimethylpolysiloxan-oil phantom accords with the reference T₂ map (mean T₂=116±1ms). The T₂* map (mean T2*=76±10ms) obtained from the EPI module is in good agreement with the MGRE reference (mean T2*=75±9ms). The T₂ and T₂* maps obtained for the tube phantom and for the brain demonstrate agreement between the data derived from 2-in-1 RARE-EPI and MSE and MGRE reference scans (Figure 3). The T₂ and T₂* maps obtained for the tube phantom and for the brain demonstrate agreement between the data derived from the 2-in-1 RARE-EPI and he MSE and MGRE reference scans (Figure 3).

Discussion and Conclusions:

This work demonstrates the feasibility of radially sampled 2-in-1 RARE-EPI which facilities recording of dual contrast weighted images (T₂ and T₂*) and simultaneous acquisition of T₂ and T₂* maps. The proposed 2-in-1 RARE-EPI hybrid technique decreases acquisition time for T₂ and T₂* mapping. To conclude, the proposed 2-in-1 RARE-EPI hybrid is not limited to brain imaging, shown in this work. Dual contrast RARE-EPI also holds the promise to meet the needs of applications dealing with bulk or physiological motion such as cardiac imaging including T₂ imaging/mapping of myocardial edema and myocardial T₂* mapping [8], as well as MRI of the eye, abdomen and liver.

Acknowledgements

The authors would like to thank Antje Els for providing assistance with the in vivo measurements.