Removal of Nuisance Signal from Sparsely Sampled H-MRSI Data Using Physics-based Spectral Bases

A novel nuisance removal method is proposed for 1H-MRSI. The method uses spectral bases generated for water and subcutaneous lipids using quantum simulation, and can perform nuisance signal removal directly from (k,t)-space data. Consequently, the proposed method is able to handle sparsely sampled MRSI data, which provides a desirable flexibility for designing accelerated 1H-MRSI data acquisition schemes. Experimental results demonstrate that the proposed method is capable of removing nuisance signals from 1H-MRSI data acquired from the brain without water and lipid suppression.