Spot-Based Global Registration for Subpixel Stitching of Single-Molecule Resolution Images for Tissue-Scale Spatial...

Area of Science:

• Molecular Biology
• Biotechnology
• Imaging Science

Background:

• Single-molecule imaging provides deep insights into cellular and tissue organization.
• Precise image stitching is crucial but challenging due to signal heterogeneity.

Purpose of the Study:

• To develop a robust method for high-resolution image alignment in tissue-scale single-molecule imaging.
• To overcome limitations of existing image-based stitching techniques.

Main Methods:

• Introduced Spot-Based Global Registration (SBGR), focusing on identified molecular spots rather than raw images.
• Developed methods for robust evaluation of registration accuracy and stitching performance.
• Implemented a spot-removal mechanism to eliminate duplicates in overlapping regions.

Main Results:

• SBGR achieves subpixel accuracy (83 ± 36 nm) with high consistency.
• The spot-based approach offers robust evaluation of stitching quality.
• Duplicate spot removal enhances information recovery from overlapping image areas.

Conclusions:

• SBGR provides a versatile and accurate solution for stitching single-molecule resolution images in spatial transcriptomics.
• The method enhances high-resolution spatial analysis of biological tissues.
• SBGR represents a significant advancement for the field of single-molecule imaging.