Meeting Banner
Abstract #2844

Water Scaling Strategy for Metabolites Quantified in MRS by Deep Learning

Yu-Long Huang1, Yi-Ru Lin1, Teng-Yi Huang2, Cheng-Wen Ko3, and Shang-Yueh Tsai4,5
1Department of Electronic and Computer Engineering, National Taiwan University of Science and Technology, Taipei, Taiwan, 2Department of Electrical Engineering, National Taiwan University of Science and Technology, Taipei, Taiwan, 3Department of Computer Science and Engineering, National Sun Yat-Sen University, Kaohsiung, Taiwan, 4Graduate Institute of Applied Physics, National ChengChi University, Taipei, Taiwan, 5Research Center of Mind, Brain and Learning, National ChengChi University, Taipei, Taiwan

Recently, it has been shown that MRS can be analyzed by a convolutional neural network (CNN) with concentrations quantified in a relative way. Here, we propose to scale in vivo MRS data according to water signal in simulated spectra and in vivo data so that CNN spectra can be scaled to institutional units for possible between subject comparison. Our results show that the quantified metabolites are at the same level as those quantified using LCModel with water scaling method but with less repeatability. A further phantom study is necessary to validate the proposed method.

This abstract and the presentation materials are available to 2020 meeting attendees and eLibrary customers only; a login is required.

Join Here