Table 1.

Signals Measured from Two 35 mm³ VOIs Placed Side By Side Inside a Homogeneous Phantom Close to the Coil

Shared excitation slice	Area of signal peak (a.u)	SNR	Signal loss
Left voxel reference	6563 ± 49	27.2 ± 1.1	–
Right voxel reference	7693 ± 48	22.8 ± 1.1	–
Left voxel MVS shared excitation	3527 ± 12	12.5 ± 0.4	46 ± 0.3%
Right voxel MVS shared excitation	4991 ± 55	10.6 ± 0.4	35 ± 1%
Shared refocusing slices
Left voxel reference	3994 ± 22	17.7 ± 0.6	–
Right voxel reference	6178 ± 55	17.7 ± 0.8	–
Left voxel MVS (1 + 7)	3607 ± 19	16.1 ± 0.6	10 ± 1%
Right voxel MVS (1 + 7)	4860 ± 50	14.7 ± 0.6	21 ± 1%
Left voxel MVS (4 + 4)	3737 ± 29	16.7 ± 0.8	6 ± 1%
Right voxel MVS (4 + 4)	5563 ± 56	16.4 ± 0.7	10 ± 1%
Left voxel MVS (4 + 4) + 12.5%	3880 ± 18	16.7 ± 0.7	3 ± 1%
Right voxel MVS (4 + 4) + 12.5%	5681 ± 44	16.5 ± 0.4	8 ± 1%

Both VOIs were acquired separately with T _R = 8 s as reference signals. For multivoxel spectroscopy (MVS), the delay between the localizations was 1 s, followed by a delay of 7 s with a shared excitation slice only and vice versa for each voxel position. The influence of sharing both refocusing pulses was tested with asymmetric (1 + 7 s) and symmetric (4 + 4 s) timing. To reduce saturation effects caused by refocusing pulses, the measurement was repeated with the refocusing pulse voltage increased by 12.5% (from 160 to 180 V).