Mass spectrometry imaging (MSI) and optical emission spectrometry imaging (OESI) are powerful label-free techniques for mapping molecular and elemental distributions in biological and chemical samples. However, these methods often suffer from artifacts such as pixel misalignment, streaking, tailing, and high-frequency noise, which compromise image quality and hinder accurate interpretation. Existing solutions predominantly rely on hardware improvements or manual curation, which are costly and not easily scalable. To address these challenges, we introduce MOESIopt, a modular computational framework that automates the restoration of suboptimal MS and OES images without requiring hardware modifications or high-resolution optical references. MOESIopt incorporates a series of tailored algorithms for low- and high-frequency feature separation, pixel alignment, destreaking, tailing correction, and adaptive filtering. The framework supports common data formats (TXT, CSV) and includes a user-friendly graphical interface for seamless integration into existing workflows. We demonstrate its effectiveness on diverse data sets from LA-DBD-OESI/MSI, IR-MALDESI, nano-DESI, and other platforms, showing significant enhancements in image clarity and structural coherence. As an open-source tool, MOESIopt offers a versatile, open-source solution to advance the reproducibility and accessibility of high-quality chemical imaging.
Mi et al. (Thu,) studied this question.