MICROSCOPIC SUSCEPTIBILITY VARIATION AND TRANSVERSE RELAXATION - THEORY AND EXPERIMENT

Citation
Rm. Weisskoff et al., MICROSCOPIC SUSCEPTIBILITY VARIATION AND TRANSVERSE RELAXATION - THEORY AND EXPERIMENT, Magnetic resonance in medicine, 31(6), 1994, pp. 601-610
Citations number
51
Categorie Soggetti
Radiology,Nuclear Medicine & Medical Imaging
ISSN journal
07403194
Volume
31
Issue
6
Year of publication
1994
Pages
601 - 610
Database
ISI
SICI code
0740-3194(1994)31:6<601:MSVATR>2.0.ZU;2-K
Abstract
Microscopic susceptibility variations invariably increase apparent tra nsverse relaxation rates. In this paper, we present comparisons betwee n Monte Carlo simulations and experiments with polystyrene microsphere s to demonstrate that this enhanced relaxation can be explained quanti tatively for both spin echo and gradient echo imaging experiments. The spheres used (1 to 30 mu), and degree of susceptibility variation (ca used by 0-12 mM Dy-DTPA) covered a wide range of biologically relevant compartment sizes and contrast agent concentrations. These results sh ow that several regimes of behavior exist, and that contrast dependenc e is quite different in these regimes. For a given susceptibility, Del ta(X), a small range of particle sizes show peak transverse relaxation . For the range of susceptibilities found in the first pass of a clini cal IV contrast agent bolus, this size range is 5 to 10 mu, or roughly capillary sized compartments. In both our simulations and experiments , smaller spheres showed quadratic relaxation versus concentration cur ves, and larger particles showed sublinear behavior. For particles cor responding to the peak relaxivity, the relaxation-concentration curves were linear. In addition, we demonstrated that increasing the diffusi on coefficient can increase, decrease, or, paradoxically, leave unaffe cted the apparent relaxation rate. The regime for which the diffusion coefficient is relatively unimportant corresponds to the region of pea k relaxivity. By using the Bloch-Torrey equation to produce scaling ru les, the specific Monte Carlo simulations were extended to more genera l cases. We use these scaling rules to demonstrate why we often find t hat susceptibility-induced relaxation rates vary approximately linearl y with concentration of injected agent.