Preliminary Study of Mapping Brain ATP and Brain pH Using Multivoxel 31P MR Spectroscopy

Abstract

Magnetic resonance (MR) spectroscopy is a valuable method for the noninvasive investigation of metabolic processes. Although brain ATP studies can be found in multivoxel 31P MR spectroscopy, previous studies of intracellular brain pH was conducted in single-voxel 31P MR spectroscopy. The purpose of this study was to explore the feasibility of mapping brain ATP and brain pH by using multivoxel 31P MR spectroscopy. Phantom studies were carried out by using a GE 3T scanner firstly. Many available sequences were tested using phantom and the 2D PRESSCSI sequence was selected because of better signal to noise ratio. TR was 1000 msec and TE 144 msec with 128 scan averages. The acquisition matrix was 16 x 16 phase encodings over a 24-cm FOV. Slice thickness was 10 mm. Then a healthy volunteer from MR research team was studied. Data were processed offline using the SAGE/IDL software. Baseline and phase corrections were performed. Multivoxel spectra and brain ATP map were analyzed. Brain pH values were calculated from the difference in chemical shifts between inorganic phosphate (Pi) and phosphocreatine (PCr) resonances. Color scaling map was generated using MatLab software. Multivoxel 31P spectra were obtained for phantom and the healthy volunteer. PCr map was obtained in phantom. At this moment, peaks of PCr were not homogeneous in phantom studies. There was noise for multivoxel 31P spectra in volunteer study. Phosphomonoester (PME) peak, Pi peak, phosphodiester (PDE) peak, PCr peak, γ{phonetic}ATP peak, αATP peak, and βATP peak can be identified. Preliminary brain ATP map and brain pH map were generated in the volunteer. It is feasible to map brain ATP and brain pH using multivoxel 31P MR spectroscopy. However, endeavors should be made to improve quality of multivoxel 31P MR spectroscopy.