Coregistration of Multimodal Imaging Mass Spectrometry and Optical Images of Tissue Sections at Single‐Cell Resolution

ABSTRACT Accurate alignment of histopathological structures in optical images and imaging mass spectrometry (IMS) datasets is crucial for providing meaningful biological interpretation. Integrating IMS with optical or other imaging modalities remains challenging due to sample‐preparation constraints, mismatched spatial resolutions, and the need for precise physical and computational alignment–especially at the single‐cell level. In this article, we discuss challenges in multimodal IMS workflows on a single tissue section and the use of laser‐etched fiducial markers to improve spatial registration. For single‐cell analyses, an open‐source deep‐learning segmentation tool was used to define cellular boundaries in tissue sections. Our results show that while consecutive‐tissue sections can align broad tissue structures, they fail to maintain cellular‐level correspondence, thereby limiting high‐resolution analyses. In contrast, same‐section multimodal coregistration using laser‐etched fiducial markers and deep‐learning segmentation enabled single‐cell molecular imaging despite reduced stain quality in the optical images. These methods provide a foundation for future advances in molecular pathology and biomarker discovery.