Premium
Application of super‐resolution reconstruction of sparse representation in mass spectrometry imaging
Author(s) -
Tang Fei,
Bi Ying,
He Jiuming,
Li Tiegang,
Abliz Zeper,
Wang Xiaohao
Publication year - 2015
Publication title -
rapid communications in mass spectrometry
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.528
H-Index - 136
eISSN - 1097-0231
pISSN - 0951-4198
DOI - 10.1002/rcm.7205
Subject(s) - interpolation (computer graphics) , chemistry , resolution (logic) , sampling (signal processing) , mass spectrometry imaging , sparse approximation , artificial intelligence , image resolution , iterative reconstruction , mass spectrometry , pattern recognition (psychology) , computer science , computer vision , image (mathematics) , chromatography , filter (signal processing)
Rationale Mass Spectrometry Imaging (MSI) is useful for analyzing biological samples directly, as a spatially resolved, label‐free technique. Here we present a method for super‐resolution reconstruction of sparse representation to improve resolution of MSI data. Methods Air Flow‐Assisted Ionization Mass Spectrometry Imaging (AFAI‐MSI) was used to acquire MSI data from ink samples, thyroid tumour samples, rat renal biopsies, and rat brain biopsy samples. Super‐resolution reconstruction of sparse representation was adopted for the collected MSI data. Results After comparison of the reconstructed high‐resolution image and the original high‐resolution image, it is found that super‐resolution reconstruction image is closer to the original high‐resolution image than the image obtained with the interpolation method, and the highest Peak Signal‐to‐Noise Ratio (PSNR) difference value is over 1.4dB. Therefore, the application of the super‐resolution reconstruction technique, based on sparse representation MSI, is feasible and effective. Conclusions The method proposed here not only improves the resolution of MSI in post‐data processing, but also acquires fewer sampling points at the same resolution, thereby greatly reducing the sampling time, with great application value for large‐volume sample MSI, high‐resolution MSI, etc. Copyright © 2015 John Wiley & Sons, Ltd.