Rapid-sequence phosphorus-31 magnetic resonance spectroscopy of the human heart using a 1.5-T clinical system

To compare a 'standard' slow phosphorus-31 magnetic resonance spectroscopy (31P-MRS) sequence with two faster sequences in phantoms and healthy volunteers using a 1.5-T clinical system.

Complete 3D localization was performed using a 2D phosphorus chemical-shift imaging sequence in combination with 30-mm axial slice-selective excitation. Two 31P-MRS rapid sequences (RS8-4: 8 x 8 phase-encoding, with an average of 4 acquisitions, and RS16-1: 16 x 16 phase-encoding, 1 acquisition) were compared with the standard sequence (StdP: 16 x 16 phase-encoding, with an average of 8 acquisitions) in phantom and healthy volunteers.

Acquisition time for the 31P-MRS procedure with StdP, RS8-4, and RS16-1 in the healthy volunteer studies ranged from 30 to 45, 3 to 5, and 3 to 5 minutes, respectively. Metabolite measurements of healthy volunteers obtained from 31P-MRS using RS8-4 correlated with values obtained using StdP (PCr r2=0.63, P<0.001; ATP r=0.41, P<0.01 and PCr/ATP ratio r2=0.25, P<0.05). There was no correlation between StdP and RS16-1 for either ATP or the PCr/ATP ratio (r2=0.03, P=0.60, and r2=0.11, p=0.26, respectively). Reproducibility (intensity of phosphorus signal) with RS16-1 was worse than that of RS8-4 or StdP.

31P-MRS using RS8-4 may be a valid diagnostic tool for patients with cardiac diseases.