CEST Analysis

Tools to register and pre-process CEST data for use with FABBER CEST.

You can use conda to install the required dependencies. To do this run:

    conda create env -n <name of conda environment> -f requirements.yml

Registering Data

Requirements:

• FSL

Running the tool

The registration pipeline utilises FSL as its backend. This tool can perform both 2D and 3D registration of CEST, B1, and VFA T1 time-series data to a reference image. To use the tool, run the following command:

    python registration.py <Your-JSON-File>.json

The JSON file used in the regisration tool is of the form:

{
    "Analysis Path": "/Path/To/Analysis/Folder",
    "Data Path": "/Path/To/Scans/Folder",
    "Scan ID": "/Name/of/Scan/Folder(s)",
    "Reference Name": "FileName",
    "CEST Names": "FileName",
    "T1 Name": "FileName",
    "B1 Anatomical Name": "FileName",
    "B1 FA Name": "FileName",
    "Offset File Names": "FileName"
}

The "Data Path" should point to the folder where all of the scan data is located. You can then list each individual scan within the "Scan ID" entry (can be a list). The "Analysis Path" is the folder you want the registered data to be stored.

MT and CEST Registration

If multiple MT/CEST files need to be registered to a single Reference image, the "CEST Names" variable will take a list. Keep in mind that "Offset File Names" also needs the same number of entries as "CEST Names".

This tool registers CEST data using the logic shown below.
If there are offsets between ±1 ppm, the the registration tool with split the stack into three separate time-series of data. The > ±1 ppm data will be registered to the S₀ image; then a co-registered image close to 1 ppm is used as the reference to MCFLIRT the remaining volumes. The three separate time-series are then combined back into one time-series for further processing. If no offsets between ±1 ppm are detected (e.g., using MT data only), then the entire time-series will be coregistered to the S₀ image.