Researchers interested in fast T₁ mapping, contrast-enhanced MRI and non-Cartesian imaging.Dynamic contrast-enhanced MRI (DCE-MRI) plays an important role in tumor imaging. The concentration of contrast agent can be quantified directly by measuring T₁ changes. However, conventional inversion recovery T₁ mapping method cannot track the dynamics of T₁ changes due to its long acquisition time. Previously, a mulitslice saturation recovery Look-Locker (MSRLL) method with Cartesian encoding was developed and validated in our laboratory for fast cardiac T₁ mapping in mice¹. In this study, the MSRLL method was modified with spiral encoding to further accelerate T₁ mapping for the application on tissues with long T₁ value, such as tumor.

A multislice saturation-recovery Look-Locker T₁ mapping method with spiral readout was modified based on the previous Cartesian MSRLL sequence¹. A spiral trajectory of 32 interleaves with zero moment compensated was designed using the minimum-time gradient method². The spiral trajectory was measured manually using an established method³ and the measured trajectory was used to reconstruct images using NUFFT⁴. Multi-slice images were acquired using the interlaced scheme to minimized cross-talk between slices. Saturation module was implemented at the beginning of each spiral interleave acquisition in k-space. T₁ maps were obtained by pixel-wise curve fitting using an in-house developed MATLAB-based software as described previously¹.

All MRI experiments were performed on a horizontal 7T Bruker scanner (Bruker Biospin Co., Billerica, MA). The spiral MSRLL method was validated in vitro on a multi-compartment phantom with manganese chloride (MnCl₂) solutions. The concentrations of MnCl₂ solutions were 0.03 mM, 0.1 mM, 0.3 mM, and 1 mM, respectively. Three slices were acquired using the spiral MSRLL method with 0.2 mm interslice gap. The following imaging parameters were used: flip angle 10°; echo time 2.09 ms; slice thickness 1 mm; number of average 1; field of view 3×3 cm²; matrix size 128×128. For each slice, 50 images that covered 5 s of the saturation recovery curve were acquired with an interval of 100 ms. Proton density (M0) images were acquired with repetition time (TR) of 2s. Cartesian MSRLL images were also acquired as a validation of spiral MSRLL method. Center 64 k-space lines were acquired in Cartesian method as described previously¹. Other acquisition parameters were the same as spiral MSRLL.

The spiral MSRLL method was then evaluated in a DCE-MRI study on mouse tumor. A 4-week old athymic nude mouse was inoculated with 1×10⁶ PC3 (ATCC® CRL1435) in flank. Imaging experiments were performed when tumor size reached 5-8 mm in diameter. An extradomain-B fibronectin (EDB-FN) targeted contrast agent, ZD2-Gd(HP-DO3A) was injected intravenously at the dose of 0.1 mmol/kg after acquiring pre-contrast images. This contrast agent preferably accumulates in tumor via targeting overexpressed EDB-FN in tumor extracellular matrix through a 7 amino-acid peptide, ZD2⁵. Post-contrast T₁ maps at 7 time points were acquired every 10 min to monitor contrast enhancement of tumor. 5 slices were acquired and other acquisition parameters were the same as the phantom study.

Fig. 1 shows T₁ maps of all three slices measured by spiral MSRLL method. There is no substantial difference in T₁ values among the slices. Total acquisition time of Look-Locker and M0 images using spiral MSRLL and Cartesian MSRLL were 3min 44s and 7min 28s, respectively. The comparison of Cartesian and spiral T₁ maps of the first slice was shown in Fig. 2a. T₁ values measured using spiral MSRLL shows a strong agreement with Cartesian method (Fig. 2b). Fig. 3 shows in vivo T₁ maps of one slice before and after contrast agent injection. The first map (Fig. 3a) was acquired before injection and the others (Fig. 3b-h) were acquired every 10 min after injection. Dynamic R₁ change is shown in Fig. 4. R₁ increased from 0.38 s^-1 to 0.54 s^-1 at 20-30 min post-injection, and decreased to 0.37 s^-1 at 70 min post-injection.The spiral MSRLL method showed a strong agreement with Cartesian method, and achieved temporal resolution of 2 min 40s with a voxel size of 0.23×0.23×1 mm³ in vivo. This fast T₁ mapping method is potentially applicable to cardiac imaging as well when combined with ECG trigger. Further acceleration can be achieved by combining with other fast imaging methods, such as compressed sensing.