Jyh-Miin Lin1, Tzu-Chao Chuang2, Wen-Chau Wu3, Hsiao-Wen Chung4, 5, Shang-Yueh Tsai6, 7
1Department of Radiology, Duke University Medical Center, Durham, NC, United States; 2Department of Electrical Engineering, National Sun Yat-sen University, Kaohsiung, Taiwan; 3Graduate Institute of Oncology, National Taiwan University, Taipei, Taiwan; 4Department of Electrical Engineering, National Taiwan University, Taipei, Taiwan; 5Institute of Biomedical Electronics and Bioinformatics, National Taiwan University, Taipei, Taiwan; 6Graduate Institute of Applied Physics, National Chengchi University, Taipei, Taiwan; 7Research Center for Mind, Brain and Learning, National Chengchi University, Taipei, Taiwan
In Vivo MR spectra without solvent suppression often incur poor spectral quality due to gradient-induced frequency modulation(FM). We invented an algorithmic method to identify and eliminate these artifacts. Advanced algebraic tool was introduced to separate symmetric peaks with opposite phases from metabolites and baseline. Based on time-domain methods, the amplitude contaminated by FM components are adjusted according to the diagonal elements of matrices which have been triangularized. Our result shows improvement of spectral quality and quantitatively lower biases. This method may be easily integrated into the processing programs, whereas the sequence and experimental design can remain similar.