Compressed sensing to accelerate magnetic resonance spectroscopic imaging: evaluation and application to 23Na-imaging of mouse hearts

摘要: Magnetic Resonance Spectroscopic Imaging (MRSI) has wide applicability for non-invasive biochemical assessment in clinical and pre-clinical applications but suffers from long scan times. Compressed sensing (CS) been successfully applied to 1H MRSI, however a detailed evaluation of CS conventional chemical shift imaging is lacking. Here we evaluate the performance accelerated specifically apply it accelerate 23Na-MRSI on mouse hearts vivo at 9.4 T. Synthetic phantom data representing simplified section across thorax were used fidelity reconstruction varying levels under-sampling, resolution signal-to-noise ratios (SNR). The amplitude signals arising within compartment, signal contamination outside compartment relative noise-free Fourier-transformed (FT) determined. Simulation results subsequently verified experimentally phantoms three vivo. reconstructed MRSI are scaled linearly absolute intensities fully-sampled FT case (R2 > 0.8, p-value < 0.001). Higher acceleration factors resulted denoising spectra, also an increased blurring boundaries, particularly lower spatial resolutions. Increasing SNR decreased cross-compartment yielded amplitudes closer data. Proof-of-concept high-resolution, 3-fold 23Na-amplitude maps murine myocardium could be obtained ~23 mins. Relative (i.e. metabolite ratios) quantification concentrations can accurately determined with up 5-fold under-sampled, CS-reconstructed MRSI. Although this work focused cardiac 23Na-MRSI, equally applicable other nuclei tissues (e.g. brain). Significant reduction time will reduce burden subject, increase scanner throughput, may open new avenues (pre-) metabolic studies.