Development and evaluation of limitation including error and reproducibility on assessment method of cerebral function using MRI

点击翻译按钮获取中文摘要

The purpose of this study is to evaluate the possibility and problem including error and reproducibility on the perfusion MRI and proton MR spectroscopy (proton MRS).On perfusion MRI, flow sensitive alternating inversion recovery (FAIR) sequence was made and installed into our clinical MRI instrument. The different inversion times on FAIR were selected and compared with the cerebral blood perfusion using SPECT. The signal ratio (SR16/8) of two images with different TIs (1600ms and 800ms) was calculated and the cerebral blood flow reserve (CFR) was evaluated by the increase in the ratio of cerebral perfusion after administration of acetazolamide in the SPECT study. A reversed linear correlation (r=0.75) was found between SR and CFR, indicating that differences of FAIR images with changes of TI will be affected by cerebral flow reserve.On proton MRS, the accuracy after curve-fitting of frequency domain data was increased by high magnetic field instrument at 3 Tesla, and the error was limited under 10%. The peak of glutamate and glutamine complex was also evaluated and the concentration of glutamate alone could be quantified with LCModel analysis on 3T MRS data. The value of SD of Glx by LCModel was under 15% and that of glutamate was under 25%.Furthermore, the sequence of MEGA-PRESS was made into 3 Tesla MRI with frequency selective 180° pulses to detect peak of gamma amino-butyric acid (GABA). By this sequence, GABA was separated from other major metabolites at 3.01 ppm and the concentration of GABA was quantified with LCModel. The SD value of GABA was under 25% and its quantified value was agreed with the literature value.The results indicates the new parameter of cerebral perfusion and the improvement of accuracy and reproducibility on proton MRS by high magnetic field at 3 Tesla.