Implications of respiratory motion for the quantification of 2D MR spectroscopic imaging data in the abdomen

and

Published under licence by IOP Publishing Ltd
, , Citation A J Schwarz and M O Leach 2000 Phys. Med. Biol. 45 2105 DOI 10.1088/0031-9155/45/8/304

0031-9155/45/8/2105

Abstract

Magnetic resonance spectroscopic imaging (MRSI) studies in the abdomen or breast are acquired in the presence of respiratory motion. This modifies the point spread function (PSF) and hence the reconstructed spectra. We evaluated the quantitative effects of both periodic and aperiodic motion on spectra localized by MRSI. Artefactual signal changes, both the modification of native to a voxel and spurious signals arising elsewhere, depend primarily upon the motion amplitude relative to the voxel dimension. A similar dependence on motion amplitude was observed for simple harmonic motion (SHM), quasi-periodic motion and random displacements. No systematic dependence upon the period or initial phase of SHM or on the array size was found. There was also no significant variation with motion direction relative to the internal and external phase-encoding directions. In measured excursion ranges of 20 breast and abdominal tumours, 70% moved ⩽ 5 mm, while 30% moved 6-23 mm. The diaphragm and fatty tissues in the gut typically moved ~ 15-20 mm. While tumour/organ excursions less than half the voxel dimension do not substantially affect native signals, the bleeding in of strong lipid signals will be problematic in 1H studies. MRSI studies in the abdomen, even of relatively well-anchored tumours, are thus likely to benefit from the addition of respiratory triggering or other motion compensation strategies.

Export citation and abstract BibTeX RIS

10.1088/0031-9155/45/8/304