MIKAIA’s Slide Alignment: An Overview

With the upcoming MIKAIA^® 3 release (scheduled for Q1 2026), MIKAIA^® will receive a new multi-omics slide alignment module, “MIKAIA^® Slide Align“. Based on a 3-phase registration process, the last one being a non-rigid deformation for highest alignment accuracy, the module lets you generate new aligned slides in SVS or OME-TIFF format. Additionally, multiple layers are tied together via a companion text file (*.multifile.csv), which can be directly opened to view the multi-layered images in MIKAIA’s multi-omic viewer.
The MIKAIA^® Slide Align module is powered by the HistokatFusion technology by Fraunhofer MEVIS.

Three use cases: Serial sections, CycIF, and multi-omics

The slide alignment module will support three main use cases:

1. Alignment of H&E and IHC serial sections (brightfield)

In a typical scenario, you will align H&E stains with one or multiple IHC stains. For each scan, an alignment version is generated, alongside a companion file that links them together.

This stack can be opened in MIKAIA^®, allowing you to toggle between the aligned brightfield layers.
You can use the various MIKAIA^® analysis apps by selecting the specific layers for analysis. For instance, use the Universal IHC Cell AI to quantify immune cells in the IHC layers, and then evalulate their colocalization with the Grid Analysis App. Or you can mask the DAB+ areas with the Mask by Color App and use these masks as training annotations for the AI Author to train an H&E AI.

This is an example of nine serial sections (H&E plus 9x IHC) aligned and displayed side-by-side in MIKAIA^®. The viewers can by navigated synchronously, with the mouse pointer mirrored from the focused viewer to the other views.

2. Alignment of Cyclic Immunofluorescence (CycIF): Create high-plex OME-TIF

In a typical scenario, you have manually stained, scanned, bleached, restained, and rescanned a tissue section using your low-plex fluorescence scanner (e.g., from Zeiss or Olympus). Let’s say you now have five cycles with four channels each: Each cycle contains a DAPI channel along with the same set of three additional fluorophores targeting different antibodies. Your goal is to align and merge these 5 4-plexes into a single high-plex OME-TIFF (e.g., DAPI + 5×3 markers = 16 plex, if you choose to discard all but one DAPI channel). Throughout this process, you can even change the channel names, colors, and order as needed.

This screenshot shows the wizard page, where all detected markers from three cyclic 4-plexes are listed and can be customized. You can modify the marker names and colors, deselect channels to discard them, and rearrange the output order .

3. Alignment of multi-omics scans (e.g., Xenium with H&E, or MALDI with H&E)

Multi-omics spatial biology promises to revolutionize discovery. A key challenge at the outset of each bioinformatic analysis is merging the various omics together to facilitate joint analysis. MIKAIA’s slide alignment module supports the alignment multi-omics scans.

Spatial transcriptomics: Align a Xenium Morphology OME-TIFF scan with an H&E brightfield scan (any scanner format) and view them togehter in the MIKAIA^® multi-omics viewer. You can then perform H&E-guided spatial transcriptomics analysis by using the AI Author to analyze the H&E morphology (e.g., finding tumors) and compare the transcriptome across different H&E-derived annotations or regions of interests.

Mass spectrometry imaging (MSI): Align a low-res/high-plex DESI or MALDI OME-TIFF image with an H&E scan. You can subsequently conduct H&E-guided MSI analysis using the AI Author to analyze the H&E morphology (e.g., locating tumor) and then investigate the lipidome or metabolome across different H&E-derived annotations, regions of interests, or cells.

MIKAIA^® Slide Align is powered by award-winning science: HistokatFusion by Fraunhofer MEVIS

Fore more details, visit the HistokatFusion website.

Ready for batch analysis

Typically, you do not just want to align one case; you need to analyze an entire cohort. Let’s say you have 100 samples, each with an H&E and five IHCs stains — 600 scans in total. To align them all, the alignment module needs to be told which six scans belong together, make up a case, and how they should be named. While this can be done manually, MIKAIA^® offers a more convenient way: You can instruct it on how to extract the case ID and stain name from the file paths. Regardless how you might have organized your files …

• \\networkshare\myslides\<case ID>\scan_<stain name>.svs
• \\networkshare\myslides\<stain name>\scan_<case ID>.svs
• \\networkshare\myslides\scan_<stain name>_<case ID>.svs

… as long as you have followed a rigid structure, you can specify the delimiters before and after the case ID and stain tokens. MIKAIA^® will then automatically group the 600 slides into 100 alignment jobs. Next, you can select which stain should serve as the reference stain (which will not be morphed), define the output file naming scheme, and set the target resolution.

Finally, all cases will be aligned automatically. Executing the three registration phases for each pair of slides only takes a few seconds. You can then preview the alignment result at a medium resolution (~5000 px per image). If the automatic alignment does not yield satisfactory results, you can enhance it by manually setting 3 landmarks in both stains.

Cropping / 1:N alignment
This is also an opportunity to optionally draw a rectangular ROI on any slide to crop it. This not only results in smaller generated aligned slides, saving disk space, but also significantly speeds up the subsequent high-resolution export.

Another scenario where cropping is beneficial is when you placed multiple specimens on a single spatial transcriptomics or proteomics scans, but created individual H&E scans for each speciment. In this case, you can simply multiply the multi-specimens scans by creating multiple shortcuts and loading them all as input scans. Then, group each copy with the corresponding single-speciment H&E scan. In the multi-specimens scan, simply draw the ROI to crop to the relevant specimen.

Finally, all cases are exported, generating new WSIs at their native resolution (unless a lower resolution has been configured). This process may take multiple minutes per case, but all cases are batch-processed, allowing for a large batch to be run overnight.

In the CycIF scenario, each case results in a single high-plex OME-TIFF. In the other use cases, multiple aligned scans are generated and an extra companion file (*.multifile.csv) links them together to a multi-omics stack. This stack is automatically added to MIKAIA’s “Slides” workspace and can be opened by double clicking it.

Funding

This MIKAIA^® extension has been kindly made possible thanks to venture capital provided by the Fraunhofer Future Foundation (Fraunhofer Zukunfsstiftung). Project: “Histology AI Author”, consortium: Fraunhofer IIS & Fraunhofer MEVIS