Development of high resolution 3D hyperpolarized carbon-13 MR molecular imaging techniques

Abstract

The goal of this project was to develop and apply techniques for T2 mapping and 3D high resolution (1.5 mm isotropic; 0.003 cm3) 13C imaging of hyperpolarized (HP) probes [1-13C]lactate, [1-13C]pyruvate, [2-13C]pyruvate, and [13C,15N2]urea in vivo. A specialized 2D bSSFP sequence was implemented on a clinical 3T scanner and used to obtain the first high resolution T2 maps of these different hyperpolarized compounds in both rats and tumor-bearing mice. These maps were first used to optimize timings for highest SNR for single time-point 3D bSSFP acquisitions with a 1.5 mm isotropic spatial resolution of normal rats. This 3D acquisition approach was extended to serial dynamic imaging with 2-fold compressed sensing acceleration without changing spatial resolution. The T2 mapping experiments yielded measurements of T2 values of greater than 1 s for all compounds within rat kidneys/vasculature and TRAMP tumors, except for [2-13C]pyruvate which was ~730 ms and ~320 ms, respectively. The high resolution 3D imaging enabled visualization the biodistribution of [1-13C]lactate, [1-13C]pyruvate, and [2-13C]pyruvate within different kidney compartments as well as in the vasculature. While the mouse anatomy is smaller, the resolution was also sufficient to image the distribution of all compounds within kidney, vasculature, and tumor. The development of the specialized 3D sequence with compressed sensing provided improved structural and functional assessments at a high (0.003 cm3) spatial and 2 s temporal resolution in vivo utilizing HP 13C substrates by exploiting their long T2 values. This 1.5 mm isotropic resolution is comparable to 1H imaging and application of this approach could be extended to future studies of uptake, metabolism, and perfusion in cancer and other disease models and may ultimately be of value for clinical imaging.

High‐resolution hyperpolarized metabolic imaging of the rat heart using k–t PCA and k–t SPARSE

“…The intrinsic spectral selectivity of steady‐state free precession has also been applied to hyperpolarized imaging, e.g. to bicarbonate imaging of the pig heart and high‐resolution (up to 1.5 mm isotropic) imaging of mouse kidneys and tumours . However, the relatively long acquisition time, on the order of 100 ms to 1 s, makes this approach impractical for rodent cardiac imaging, where the RR interval is around 150 ms and cardiac motion would distort the images substantially.…”

“…Parallel imaging has been used with acceleration factors of up to 3.75, requiring, however, multi‐channel array coils, which are not widely available on small‐animal imaging systems. As hyperpolarized imaging lacks background signal, the resulting images are inherently sparse and therefore attractive for compressed sensing reconstruction, where acceleration factors of up to 3.8 have been reported . Spatiotemporal correlations have been exploited using k – t principal component analysis (PCA) to accelerate dynamic acquisition by an effective factor of 3.5 .…”

Noninvasive Interrogation of Cancer Metabolism with Hyperpolarized 13C MRI

“…However, the steady-state free precession excitation bandwidths may be narrower than in vivo linewidths because of B 0 inhomogeneity such that acquisitions from multiple pulse trains, offset in frequency, must be combined (14) to achieve sufficient spectral coverage at the expense of temporal resolution. Higher temporal resolution can be achieved with compressed sensing acceleration (15).…”

“…If labeling at a position with directly attached protons is desired, substituting these protons with deuterium can increase T 1 values. Directly attached 14 N (99.6% natural abundance) also reduces T 1 values, which can be increased via 15 N substitution (27).…”

The Need and Initial Practice of Parallel Imaging and Compressed Sensing in Hyperpolarized 13C MRI in vivo

“…The results can be compared to the current compressed sensing implementation [40] in terms of SNR, spatial resolution, and dynamic acquisition capabilities. The 3D version of the sequence (two phase encoding dimensions), with an extension to the temporal dimension, would be a good candidate for parallel imaging due to the potentially large matrix sizes, especially as this approach is possibly transferred from preclinical to clinical imaging.…”