In vivo phosphorus magnetic resonance spectroscopic imaging of the whole human liver at 7 T using a phosphorus whole-body transmit coil and 16-channel receive array: Repeatability and effects of principal component analysis-based denoising

Abstract

Quantitative three-dimensional (3D) imaging of phosphorus (31P) metabolites is potentially a promising technique with which to assess the progression of liver disease and monitor therapy response. However, 31P magnetic resonance spectroscopy has a low sensitivity and commonly used 31P surface coils do not provide full coverage of the liver. This study aimed to overcome these limitations by using a 31P whole-body transmit coil in combination with a 16-channel 31P receive array at 7 T. Using this setup, we determined the repeatability of whole-liver 31P magnetic resonance spectroscopic imaging (31P MRSI) in healthy subjects and assessed the effects of principal component analysis (PCA)-based denoising on the repeatability parameters. In addition, spatial variations of 31P metabolites within the liver were analyzed. 3D 31P MRSI data of the liver were acquired with a nominal voxel size of 20 mm isotropic in 10 healthy volunteers twice on the same day. Data were reconstructed without denoising, and with PCA-based denoising before or after channel combination. From the test–retest data, repeatability parameters for metabolite level quantification were determined for 12 31P metabolite signals. On average, 31P MR spectra from 100 ± 25 voxels in the liver were analyzed. Only voxels with contamination from skeletal muscle or the gall bladder were excluded and no voxels were discarded based on (low) signal-to-noise ratio (SNR). Repeatability for most quantified 31P metabolite levels in the liver was good to excellent, with an intrasubject variability below 10%. PCA-based denoising increased the SNR ~ 3-fold, but did not improve the repeatability for mean liver 31P metabolite quantification with the fitting constraints used. Significant spatial heterogeneity of various 31P metabolite levels within the liver was observed, with marked differences for the phosphomonoester and phosphodiester metabolites between the left and right lobe. In conclusion, using a 31P whole-body transmit coil in combination with a 16-channel 31P receive array at 7 T allowed 31P MRSI acquisitions with full liver coverage and good to excellent repeatability.