1H-MRS internal thermometry in test-objects (phantoms) to within 0.1 K for quality assurance in long-term quantitative MR studies

Many magnetic resonance test-object properties are temperature-dependent, with typical temperature coefficients of similar to 2-3% K-1. Therefore, to achieve consistent quality assurance measurements to within 1%, test object temperatures should ideally be known to within 0.3 K. Proton magnetic resonance spectroscopy has previously been used to estimate accurately absolute tissue temperature in vivo, based on the linear temperature dependence of the chemical shift difference between water and temperature-stable reference metabolites such as N-acetylaspartate. In this study, this method of 'internal thermometry' in quality assurance test-objects was investigated, and in particular the value of sodium 3-(trimethylsilyl)propane-1-sulfonate (DSS) as a chemical shift reference was demonstrated. The relationship between the DSS-water chemical shift difference (sigma, expressed in ppm) and temperature tau (in K) was shown to be tau = 764.55 (+/- 5.05) - 97.72 (+/- 1.05) sigma (286 <= tau <= 309 K). Internal thermometry in MRI test-objects is feasible and straightforward, using readily available H-1-MRS pulse sequences and standard spectroscopy evaluation packages, with a minimum detectable temperature difference of 100 (+/- 20) mK. Copyright (C) 2006 John Wiley & Sons, Ltd.