Spatial analyses of tumour biopsies offer profound insights into the dynamics of tumour growth and response to treatment. Yet, the limitation of markers detectable on a single tissue section constrains these analyses. Traditional methods of staining and imaging multiple markers often result in spatial misalignment due to variations in tissue placement, deformations, and changes in physical structure across serial sections. This misalignment restricts the number of markers that can be analysed concurrently, thwarting comprehensive insights. Image registration, the process of aligning images within a common coordinate system, emerges as a solution to this challenge. By aligning histology images, it offers the potential to overcome misalignment issues and unlock the full power of spatial analysis. However, the unique characteristics of histology images, including colour intensity variations, staining differences, and tissue deformations, present difficult challenges.

In response to these challenges, a revolutionary software called Virtual Alignment of pathoLogy Image Series (VALIS) has been developed. This fully automated, flexible, and robust software offers a comprehensive solution to WSI registration. Unlike previous methods that required a reference image, VALIS employs a groupwise registration approach that aligns images serially based on their similarity. This eliminates the need for reference image selection, a critical factor in the success of registration. VALIS boasts a range of features that enhance its usability and applicability. It accommodates brightfield and immunofluorescence (IF) images, accommodating the increasing demand for multiplexed imaging. While VALIS offers great capabilities, it is not without limitations. Although it has been tested on thousands of images and proven highly successful, there may still be cases where registration fails. Additionally, its performance is constrained by CPU processing, which may lead to time-consuming tasks for very large images.

VALIS paves the way for constructing highly multiplexed images by combining staining cycles and image registration. By selecting staining sequences strategically, it is possible to enhance the number of markers that can be analysed, thus maximising the insights gained from spatial analyses.

Virtual alignment of pathology image series for multi-gigapixel whole slide images

Chandler D. Gatenbee, Ann-Marie Baker, Sandhya Prabhakaran, Ottilie Swinyard, Robbert J. C. Slebos, Gunjan Mandal, Eoghan Mulholland, Noemi Andor, Andriy Marusyk, Simon Leedham, Jose R. Conejo-Garcia, Christine H. Chung, Mark Robertson-Tessi, Trevor A. Graham & Alexander R. A. Anderson