Mass spectrum data processing based on compressed sensing recognition and sparse difference recovery

A new compressed sensing (CS) framework is presented for intelligent mass spectrum data processing in this paper. MS sensing data is used to realize the prior MS analysis through compressed sensing recognition (CSR) method. Then, based on the CSR prior knowledge, we propose the concept of sparse difference (SD) to accomplish high quality CS recovery for high dimensional MS data. The effectiveness and feasibility of the proposed method is validated by numerical experiments.