# IntegrAlign: a comprehensive tool for multi-immunofluorescence panel integration through image alignment

## TL;DR

IntegrAlign is a new tool that aligns multiple immunofluorescence images to study tissue structures and cell interactions with high precision.

## Contribution

IntegrAlign introduces a precise two-step image alignment method for multi-immunofluorescence panels, improving spatial analysis accuracy.

## Key findings

- IntegrAlign achieves alignment accuracy below 6 µm, outperforming existing methods.
- The tool supports combining extended phenotypic data from serial whole slides for spatial analysis.
- It is open-source and validated on both simulated and real datasets.

## Abstract

Tyramide-based multiplex-immunofluorescence (mIF) enables the simultaneous analysis of up to seven protein markers on a whole slide, providing a comprehensive approach to study the tumor microenvironment. Integrating multiple mIF panels through image alignment of serial slide significantly expands the number of cell populations analysed in a single space. IntegrAlign was developed to optimize this integration on serial whole slides, enhancing the value and applicability of -mIF for comprehensive spatial analyses and enabling biomarker discovery at scale.

IntegrAlign, leveraging the SimpleITK toolkit, applies a two-step alignment using rigid and B-spline transformations to integrate serial mIF whole slides. Validation on simulated and real datasets demonstrated alignment accuracy below the diameter of a cell nucleus (∼6 µm), outperforming existing methods. This precision enhances spatial analyses by combining extended phenotypic data, supporting novel insights into tissue architecture and cellular interactions.

IntegrAlign is open-source, implemented in Python, and freely available under the MIT license at https://github.com/CAUXlab/IntegrAlign.