| Summary: | 博士 === 國立臺灣科技大學 === 電機工程系 === 100 === In MRI, some specific scan parameters such as inversion time or RF frequency offset have to be determined before applying clinical protocols. In general, the MR operators or doctors would utilize the developed scouting protocol and then select the proper imaging parameter according to the set of images manually. However, the additional scouting scan would diminish the fluency of clinical routine. In this thesis, we proposed a framework of real-time feedback system that could automatically and rapidly optimize the desired parameter. Two types of applications have been accomplished based on this real-time feedback system. The first application was to implement the optimization of compensative gradient moment for z-shim method in gradient-echo echo-planar-imaging (GE-EPI). In this type, the pre-scan and target sequences were identical. After about 20 iterations for evaluating optimal negative and positive z-shim gradient moments, the following z-shimmed GE-EPI for functional MRI (fMRI) study would contain images alternatively according to the inherited values. The second application is the calibration of RF frequency to suppress the dark banding artifact.in steady-state free precession (bSSFP) cardiac imaging at 3.0 T. The real-time feedback system combined two sequences: transition-band low flip-angle bSSFP (bSSSFP-L) for frequency calibration and conventional cine bSSFP. The optimal frequency obtained by real-time feedback calibration significantly reduced dark-band artifacts in cine bSSFP images (P < .01). The period of calibration and conventional protocol could be implemented within one breadth-hold. In conclusion, the proposed real-time feedback system is rapid and fully automatic and can thus serve as a pre-adjustment module in various MR researches which are in need of parameter calibration.
|
|---|
