Attention: Data Proxy will be migrated from SWIFT to S3 storage at Friday, the 9th of May 2025 starting from 9pm CEST (my timezone). For more details, please join the rocket chat channel https://chat.ebrains.eu/channel/data-proxy-user-group


3. Image registration to reference atlas using QuickNII and VisuAlign

Version 21.2 by puchades on 2020/04/07 16:12

BICCN_QNII_figure.png

The QUINT workflow allows you to register series of histological section images from the brain to a 3D reference atlas such as the Allen Mouse Brain Atlas or the Waxholm Space atlas of the rat brain. This is a two-step process:

  1. QuickNII guides you through an interactive affine alignment, including propagation of alignment settings across the section series.
  2. VisuAlign lets you fine-tune your alignment using non-linear adjustments.

Further steps of the QUINT workflow are compatible with the atlas map output of both QuickNII and VisuAlign (the VisuAlign step is optional).

Preparation of the image series

Before you get started with image registration, make sure your images are the right size, and files are named according to the QUINT naming convention. See 2. Image pre-processing with Nutil Transform for details.

Generate your image descriptor file with FileBuilder

Use the small program “FileBuilder.bat” provided with QuickNII. A new window will open, and ask for the folder where your images are located. Point to the correct folder, mark all image files, and click OK. An XML file is generated. Always save this file in the same folder as the resized images.

Open QuickNII and load your images

Open the QuickNII program from the .exe file. Once the program opens, click the Manage data button and load your XML file.

A detailed QuickNII user manual as well as a demo dataset can be found on https://www.nitrc.org/projects/quicknii/

Anchoring procedure

The alignment procedure consists of two main tasks:

  1. Determine the dorsoventral and mediolateral cutting angle of the section series, apply same angles for all slices.
  2. Adjust the atlas slices generated along these angles to match individual sections.

A) To determine the cutting angles:

-Open one image in the beginning of the series and use anatomical landmarks to find the approximate anteroposterior position of the section. Select the atlas modality to be overlayed on the image (MRI, Atlas, etc..)

-Use the transparency slider continuously to determine how well the atlas fits the section. Determine the sectioning angles (dorso-ventral and medio-lateral) by tilting the atlas and adjust the atlas borders.

-Save the positions for this section for now, they can be adjust later.

-Open another image, located at the other end of the series, and repeat the same procedure as above, i.e. find the approximate anteroposterior position of the section and test how well the cutting angles fit this section. Save the positions.

-Explore a few sections in order to find the angles that fit most sections in order to achieve a global anchoring.

-By having the same cutting angles set for two sections (i.e. ML=+1; DV= -4), QuickNII will apply them to all the sections located between them. The antero-posterior positions of all the images are also estimated based on the numbering.

-Review the atlas to section match for all the sections, adjusting the atlas borders as necessary, and save.

-Export customised atlas maps corresponding to your series by clicking on the “export Slices” button. The exported .flat files are used for the quantification by Nutil Quantifier.

Nonlinear adjustment with VisuAlign

VisuAlign_illustration.png

 A detailed VisuAlign user manual can be found on https://www.nitrc.org/docman/?group_id=1426

The basic steps are:

-Use the same folder as QuickNII with your png images and series descriptor (XML file). You will need the JSON version of the descriptor (it is generated by QuickNII by clicking “Save JSON”).

-Open tool by clicking on “VisuAlign.bat” file. It loads and saves JSON files created by QuickNII.

-Add markers and drag them to move the atlas borders. Start with the borders of the tissue and then adjust internal landmarks.

NB: keep the number of markers to a minimum.

-The outline mode is toggled by pulling the opacity slider to the far right (as seen in the screenshot). A color picker to the right of it becomes active for changing outline colour.

-“Debug mode” displays triangles representing the deformation field, it is enabled from the View menu (and also enables a colour picker for changing the colour of the triangles).

- Use the “Save As”  to save your deformations as a new JSON file.