Datensatz

Zusammenfassung

Purpose/Introduction: The biomolecule creatine is involved in energy metabolism in the muscle [1]. Mapping of phosphocreatine (PCr) and pH in vivo can be achieved using 31P spectroscopic imaging (SI) while Chemical Exchange Saturation Transfer (CEST) enables the detection of creatine [2]. The AREX method [3] in the pulsed case [4] enables the determination of relative concentration fB and exchange rate kBA by means of the X-plot approach [4]. We show that this quantitative method correlates well with the observed changes of 31P peak areas and pH obtained by in vivo 31P-SI in the human calf muscle at rest and during exercise. Subjects and Methods: The apparent exchange-dependent relaxation rate in the case of saturation by a train of Gaussian–shaped rf pulses is given in steady state by4: AREXGaussian = DCfBkBAc1x1 2 / (x1 2 + kBA(kBA + R2B)c2 2 ). By fitting the linear relation (1/AREXGaussian = m•1/x1 2 + n) relative concentration and exchange rate can be calculated: fB = (DCnc1(-R2B 2 + H(R2B 2 /4 + m/ (nc2 2 ))))-1 and kBA = - R2B 2 + H(R2B 2 /4 + m/(nc2 2 )). Using the exchange rate the pH value can be derived [5]. Imaging: Z–spectra were obtained by centric–reordered 2D–GRE–CEST MRI (resolution = 1.4 9 1.4 9 5 mm3) implemented on a 7–T whole–body scanner (MAGNETOM; Siemens, Germany) using a 28–ch Rx/1–ch Tx 1H knee coil. For saturation 40 Gaussian–shaped rf pulses (tp = 0.1 s, DC = 50 ) with four different B1 amplitudes were applied. 31P-NMR spectra were obtained using 2D echo-planar spectroscopic imaging (EPSI) and a double-resonant 31P-1H volume coil (Rapid Biomed, Rimpar, Germany). The 31P-EPSI6 parameters were: resolution = 10 9 10 9 40 mm3, TR = 300 ms, 31P-1H NOE preparation 120 ms; tacq: 60 s. The exercise was conducted using an MR-compatible foot pedal during the CEST saturation. The same time pattern was employed in the 31P-SI measurements. Results: The equations for fB and kBA enable pixel–wise quantitative evaluation of the CEST data. The generated maps of changes between exercise and rest are shown in Figure 1 for 1H–CEST (A,C) and 31PSI (B,D). A pixel–by–pixel analysis using a scatter plot (Figure 2) reveals a strong correlation between the changes in concentrations (R2 = 0.99) and pH (R2 = 0.97). The pH values show a correlation close to 1. Discussion/Conclusion: Using AREX and the X-plot method in the pulsed case [4] allowed the determination of relative concentration fB and pH in vivo in the human leg muscle. Comparison with 31P-EPSI revealed other contributions than the intracellular creatine to the 1.9 ppm resonance, but supports the pH mapping via CEST. Altogether, the presented approach forms the first full quantitative pulsed CEST imaging applicable at a whole–body MR scanner.