Synopsis

An EPI acquisition can efficiently acquire data, allowing multiple images to be acquired before the transverse signal decays due to T2*. These images (acquired at different echo times) can be reconstructed into images of individual species, as with Dixon or IDEAL. This approach can be used to efficiently acquire hyperpolarized data by limiting the number of excitations to preserve longitudinal magnetization for subsequent time points or enabling the use of larger flip angles for higher SNR. In this abstract, we demonstrate the feasibility of this approach with thermal phantoms (acetate, alanine, and urea) and with hyperpolarized phantoms (pyruvate and urea).

Introduction

Methods

All experiments were performed on a 3T preclinical BrukerBiospec scanner using a dual-tuned transmit/receive ¹H/¹³C birdcage coil. Images were acquired with a T₂^* EPI sequence: multiple EPI readouts with different echo times are acquired following a single RF excitation (Figure 1). A 16x16 matrix is acquired with a 32x32-cm FOV. A 2DEPSI sequence⁶ was used to determine the ratio of chemical species in each voxel for comparison.

Thermal Experiment Phantoms composed of various mixtures of ¹³C-enriched chemicals were prepared for testing the method. Phantom 1 consisted of [¹³C]urea and [1-¹³C]acetate; and Phantom 2 consisted of [¹³C]urea, [1-¹³C]alanine, and [1-¹³C]acetate. For thermally polarized phantoms, the following acquisition parameters were used: TR=1000ms, TE₁=11.5ms, ∆TE=6.3ms, and 300 averages for a total scan time of 5min.

Hyperpolarized experiment To test the performance of the reconstruction in a hyperpolarized setting, two phantoms were imaged: Phantom 1 consisted of [1-¹³C]pyruvate. Phantom 2 consisted of [1-¹³C]pyruvate and [¹³C]urea. For hyperpolarized phantoms, the following acquisition parameters were used: TR=200 ms, TE₁=11.5ms, ∆TE=4.7ms, and 150 time points for a total scan time of 30s.

Reconstruction Complex DICOM images were processed offline in MATLAB with an iterative reconstruction⁸. The chemical shifts of the phantoms were determined using a non-localized spectroscopy sequence.

Results

Thermal experiment 2DEPSI spectra are used to determine the ratio of metabolites in each phantom (Figure 2). Individual EPI images are shown as well as the reconstructed metabolite images. As expected, the relative amplitude of the phantoms changes with TE as the different species destructively interfere due to different resonance frequencies. In the metabolite images, we see that Phantom 1 only appears in the urea and acetate images, while Phantom 2 appears in all three metabolite images. This demonstrates that the reconstruction correctly decomposed into the corresponding metabolite images.

Polarized experiment The intensity of the HP pyruvate is so large, that it is difficult to visualize the difference in signals due to the dynamic range. Instead, we present the average signal in ROIs in the metabolite images compared to the area under the curves of the spectra normalized to the pyruvate signal. The values of the spectroscopic data (second number) are lower due to T₁ decay and RF excitations between the acquisitions, but show a similar trend.

Hydrate Urea

Phantom1: 12.7/2.6% 6.3/3.7%

Phantom2: 14.7/4.5% 10.4/5.3%

Discussion

A challenge for Dixon methods is identifying the species when only one is present in a voxel, a particular challenge for our phantom experiment where there is physical separation of the phantoms, an additional challenge for region-growing algorithms. The HP pyruvate signal resulted in an essentially single-species phantom. However, in an in vivo setting, the pyruvate and lactate signals will be on the same order and it will be easier to generate a contiguous field map, which will improve the robustness of the reconstruction.

The number of echoes that can be acquired, and thus the number of metabolites that can be reconstructed, is limited by T₂^* and resolution. Since all echoes are acquired after a single excitation, the echo spacing is restricted to values larger than the acquisition time. However, the overall scan time is shorter than a spatial-spectral or standard IDEAL acquisition because fewer repetitions are needed to acquire multiple images. With our resolution, we are still able to acquire 5 echoes in under 50ms. While we use an EPI readout, this method can be used by other single-shot trajectories, such as spiral.

Conclusion

Acknowledgements

References

1. Vander Heiden MG, Cantley LC, Thompson CB. Understanding the Warburg Effect: The Metabolic Requirements of Cell Proliferation. Science. 2009;324(5930):1029-1033. doi:10.1126/science.1160809

2. Nelson SJ, Kurhanewicz J, Vigneron DB, et al. Metabolic imaging of patients with prostate cancer using hyperpolarized [1-13C] pyruvate. Sci Transl Med. 2013;5(198):198ra108–198ra108.

3. Lau AZ, Chen AP, Ghugre NR, et al. Rapid multislice imaging of hyperpolarized ¹³C pyruvate and bicarbonate in the heart. Magn Reson Med. 2010;64(5):1323-1331. doi:10.1002/mrm.22525

4. Lau JYC, Geraghty BJ, Chen AP, Cunningham CH. Improved tolerance to off-resonance in spectral-spatial EPI of hyperpolarized [1-13C]pyruvate and metabolites. Magn Reson Med. 2018;80(3):925-934. doi:10.1002/mrm.27086

5. Wiesinger F, Weidl E, Menzel MI, et al. IDEAL spiral CSI for dynamic metabolic MR imaging of hyperpolarized [1-¹³ C]pyruvate. Magn Reson Med. 2012;68(1):8-16. doi:10.1002/mrm.23212 6. Chen AP, Cunningham CH, Ozturk‐Isik E, et al. High-speed 3T MR spectroscopic imaging of prostate with flyback echo-planar encoding. J Magn Reson Imaging. 2007;25(6):1288-1292. doi:10.1002/jmri.20916

7. Morze C von, Larson PEZ, Hu S, et al. Imaging of blood flow using hyperpolarized [¹³C]Urea in preclinical cancer models. J Magn Reson Imaging. 2011;33(3):692-697. doi:10.1002/jmri.22484

8. Reeder SB, Wen Z, Yu H, et al. Multicoil Dixon chemical species separation with an iterative least-squares estimation method. Magn Reson Med. 2004;51(1):35-45.