Hsiu-Fen Lin1, Cheng-Wen Ko1, Shang-Yueh Tsai2, 3, Ping-Hong Lai4, 5, Thomas Lange6
1Dept. of Computer Science and Engineering, National Sun Yat-sen University, Kaohsiung, Taiwan; 2Graduate Institute of Applied Physics, National Chengchi University, Taipei, Taiwan; 3Research Center for Mind, Brain and Learning, National Chengchi University, Taipei, Taiwan; 4Dept. of Radiology, Veterans General Hospital-Kaohsiung, Kaohsiung, Taiwan; 5School of Medicine, National Yang-Ming University, Taipei, Taiwan; 6Department of Radiology, University Medical Center Freiburg, Freiburg, Germany
Amino acids such as valine, isoleucine and leucine have been recognized as key metabolites to categorize brain abscesses. In this study, we investigated the potential underestimation of amino acids caused by basis spectra for abscess patients. Short TE spectra were analyzed by LCModel using basis set of AA and that of valine, isoleucine and leucine respectively. Cramer-Rao lower bounds were used as the criteria to identify the presence of amino acids in MRS. Our result suggests that using individual model spectra as basis sets would achieve better fitting than using synthetic multiple model spectra for J-coupled metabolites.