Data from: Precise MRI-histology coregistration of paraffin-embedded tissue with blockface imaging
Data files
May 16, 2026 version files 55.06 GB
-
2mm_data_final_lowerRes.tar.gz
55.06 GB
-
README.md
5.27 KB
Abstract
Magnetic resonance imaging (MRI) provides 3D spatial information on tissue, yet it lacks at the molecular level. In contrast, histology provides cellular and molecular information, but it lacks the 3D spatial context and direct in vivo translation. Coregistering the two is key for the 3D-embedding of histological details, validating pathological MRI findings, and finding quantitative imaging biomarkers of neurodegenerative diseases. However, coregistration is challenging due to non-linear distortions of the tissue from histological processing and sectioning leading to microscopic and macroscopic nonlinear 3D deformations between specimen MRI and stained histology sections. To address this, we developed a novel pipeline integrating robust optical approaches with innovative 2D and 3D registration algorithms to achieve precise volumetric alignment of specimen MRI data with histological images. On a variety of brain tissue specimens from distinct anatomical regions and across multiple species, our methodology generated blockface volumes with minimal distortion and artifacts. Using these blockface volumes as an intermediary, we achieve a precise alignment between MRI and histology slides, yielding registration results with an overlapping Dice score of ~90% for whole tissue alignment between MRI and blockface volumes, and >95% for 2D MRI-histology registration. This correlative MRI-histology pipeline with robust 2D and 3D coregistration methods promises to enhance our understanding of neurodegenerative diseases and aid the development of MRI-based disease biomarkers.
Dryad DOI: https://doi.org/10.5061/dryad.cnp5hqcgh
This repository contains example data accompanying the our paper in Imaging Neuroscience 2025: Achieving Precise MRI–Histology Coregistration with Blockface Imaging
The corresponding codebase for generating these outputs is available at:
https://code.stanford.edu/zeinehlab/BBQ
Overview of example dataset
File: 2mm_data_final_lowerRes.tar.gz
An example dataset (one human 2mm cortex specimen) containing all required inputs, intermediate files, and outputs produced by the MR–histology coregistration pipeline.
The dataset is organized into the following directories:
Directory and file descriptions
MR/
Contains the original MRI data used as input to the pipeline.
File types:
.nii.gz— NIfTI-1 compressed MRI volumes
Example files:
7_MGE_iso_BW125_10TEs_ES...mean_2mmNo1_masked.nii.gz
Software to open:
- FSL (free, https://fsl.fmrib.ox.ac.uk)
- ITK-SNAP (free, https://www.itksnap.org)
- 3D Slicer (free, https://www.slicer.org)
scannedSlides/
Contains all histology-related data and MR–histology correspondence outputs.
histo/
Scanned histology slides.
File types:
.svs— whole-slide histology images
Naming convention:
- Files are named according to their corresponding blockface slice number
(e.g.,histo-423.svs,histo-424.svs).
Software to open:
- QuPath (free, https://qupath.github.io)
- OpenSlide + compatible viewers (free, https://openslide.org)
mr_img_spline/
MR images resampled and warped to histology space using spline-based interpolation.
File types:
.tif— 2D TIFF images representing MR correspondence for each histology slide
Software to open:
- ImageJ / Fiji (free, https://fiji.sc)
- Napari (free, https://napari.org)
Tirl3D_spline/
Standardized 2D MR–histology registration outputs for each histology slide.
These are final, standardized outputs of the pipeline representing slice-wise 2D registration between MR and histology.
File types:
- Folders per slide (e.g.,
mr-tirl3d-442-b2h/) - Registration parameter files and transformation results (text-based)
Context:
Each folder contains the transformation parameters that map MR slices to corresponding histology slides using the TIRL (Toolkit for Image Registration and Localization) framework.
Software to use:
- TIRL (free, academic, https://github.com/zeinehlab/tirl)
- Standard text editors for inspection
blockall/
Blockface imaging data used for 3D reconstruction and perspective correction.
File types:
.bmp— blockface images.nii/.nii.gz— 3D blockface volumes.txt— metadata and configuration files, intermediate output generated by BBQ pipeline.mat— MATLAB-format auxiliary files, intermediate output generated by BBQ pipeline
Example files:
blockface-423.bmp,blockface-424.bmp, …grid.bmp— calibration grid for perspective correctiontranslated...- all intermediate processed outputs generated by the algorithmExcludeTheseSlices.txt- slice number that need to be excluded from the analysis
Software to open:
- ImageJ / Fiji (BMP)
- FSL / ITK-SNAP / 3D Slicer (NIfTI)
- GNU Octave (free MATLAB-compatible alternative, https://www.gnu.org/software/octave)
Tirl3D/
Standardized 3D registration outputs between MRI and the reconstructed blockface volume.
This directory contains the final 3D alignment results and intermediate registration steps.
File types:
.nii.gz— fixed and moving volumes at different registration stages.img.chain— transformation chains defining composite mappings.log/.txt— parameter logs and registration diagnostics
Key files:
fixed.nii.gz,moving.nii.gzfixed2_rigid,fixed4_affine,fixed5_nonlinearsource_to_target.img.chain,target_to_source.img.chain
Software to use:
- TIRL (free, academic)
- FSL / ITK-based tools for visualization
- Standard text editors for logs
mri_to_blockface_2mmblock2.yml
Configuration files for the 3D MR-blockface registration
Human subjects data
We procured de-identified human brain specimens from distinct anatomical sites, including human cortex, under Stanford IRB protocol nr 33727. Informed consent was obtained for all postmortem tissue samples used in this study. Specimens were fully anonymized prior to transfer to the research team; no direct identifiers or re-identifiable information were available to the investigators. The human cortex block was dissected into multiple fragments of varying thickness, from which 2-mm slices were selected.