MIRACL in a nutshell

MIRACL (Multi-modal Image Registration And Connectivity anaLysis) is a general-purpose, open-source pipeline for automated:

  1. Registration of cleared and imaging data (ex. LSFM and MRI) to atlases (ex. Allen Reference Atlas)

  2. 3D Segmentation and feature extraction of cleared data

  3. Tract-specific or network-level connectivity analysis

  4. Statistical analysis of cleared and imaging data

  5. Comparison of dMRI/tractography, virus tracing, and connectivity atlases

  6. Atlas generation and Label manipulation

Program structure

MIRACL is structured into Modules and Workflows.

Modules

The pipeline is comprised of different Modules depending on their respective functionality. Functions for each module are grouped together:

Module

Functionality

connect

Connectivity

conv

Conversion (Input/Output)

reg

Registration

seg

Segmentation

lbls

Labels

utilfn

Utilities

sta

Structure Tensor Analysis

stats

Statistics

An example of using a module would be to run the clar_allen function which performs a CLARITY whole-brain registration to Allen Atlas on a nifti image (down-sampled by a factor of five):

$ miracl reg clar_allen -i niftis/SHIELD_05x_down_autoflor_chan.nii.gz -o ARI -m combined -b 1

The above command uses the -i flag to select the nifti file, -o to specify the orientation of the image, -m to register to both hemispheres and -b to include the olfactory bulb.

Workflows

The workflow (flow) module combines multiple functions from the above modules for ease of use to perform a desired task.

For example, a standard reg/seg analysis could look like this:

First perform registration of whole-brain CLARITY data to ARA:

$ miracl flow reg_clar -h

Then perform segmentation and feature extraction of full resolution CLARITY data:

$ miracl flow seg -h

Or structure tensor analysis:

$ miracl flow sta -h