This work presents a novel spectral model for spectral quantification, which represents each spectral component using a subspace instead of a parametric basis function with unknown parameters. The proposed model enables efficient and effective incorporation of both spectral and spatial prior information to improve the quantification performance. The proposed method is validated using both simulation and experimental data, demonstrating superior performance to existing methods using parametric spectral bases. This method is expected to be useful for processing noisy MRSI data.
This abstract and the presentation materials are available to members only; a login is required.