Abstract 2078: VALIS: Virtual Alignment of pathoLogy Image Series for multi-gigapixel whole slide images

Abstract

Abstract Interest in spatial omics is on the rise, but generation of highly multiplexed images used in many spatial analyses remains challenging, due to cost, expertise, methodical constraints, and/or access to technology. An alternative to performing highly multiplexed staining is to register collections of whole slide images (WSI), creating a collection of aligned images that can undergo spatial analyses. However, registration of WSI is two part problem, with the first being the alignment itself, and the second being the application of the transformations to huge multi-gigapixel images. To address both challenges, we have developed the Virtual Alignment of pathoLogy Image Series (VALIS) software, which enables one to rapidly and easily generate highly multiplexed images by aligning (registering) any number of multi-gigapixel whole slide images (WSI) stained using immunohistochemistry (IHC) and/or immunofluorescence (IF). Benchmarking using the publicly available 2019 ANHIR and 2022 ACROBAT Grand Challenge datasets indicates that VALIS provides state of the art accuracy, being one of the most accurate publicly available methods in the ANHIR challenge, and the most accurate opensource method in the ACROBAT challenge. VALIS is able to read, register, and save multi-gigapixel images as ome.tiff, thereby addressing the second challenge. In addition to the benchmarking datasets, the generalizability of VALIS has been tested with 273 IHC samples and 340 IF samples, each of which contained between 2-69 images per sample. In total, VALIS has therefore been tested with 5,138 images. The registered WSI tend to have low error and are completed within a matter of minutes. VALIS is written in Python, requires only few lines of code for execution, is readily available and fully documented. VALIS therefore provides a free, opensource, flexible, scalable, robust, and easy to use pipeline for rigid and non-rigid registration of multi-gigapixel WSI, facilitating spatial analyses of prospective and existing datasets,breathing new life into the countless collections of brightfield and immunofluorescence images. Citation Format: Chandler Dean Gatenbee, Ann-Marie Baker, Sandhya Prabhakaran, Mark Robertson-Tessi, Trevor Graham, Alexander R. Anderson. VALIS: Virtual Alignment of pathoLogy Image Series for multi-gigapixel whole slide images [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2023; Part 1 (Regular and Invited Abstracts); 2023 Apr 14-19; Orlando, FL. Philadelphia (PA): AACR; Cancer Res 2023;83(7_Suppl):Abstract nr 2078.