In neuroconductor/malf.templates: Template Images for Multi-Atlas Label Fusion ('MALF')

Data Disclaimer

This package wraps the templates with the cerebellum from the MICCAI 2012 Grand Challenge and Workshop on Multi-Atlas Labeling http://masiweb.vuse.vanderbilt.edu/workshop2012/index.php/Main_Page.

The challenge details are located at http://masiweb.vuse.vanderbilt.edu/workshop2012/index.php/Challenge_Details

The data were be released under the Creative Commons Attribution-NonCommercial (CC BY-NC) with no end date. Users should credit the MRI scans as originating from the OASIS project and the labeled data as "provided by Neuromorphometrics, Inc. (http://Neuromorphometrics.com/) under academic subscription". These references should be included in all workshop and final publications.

A summary of this message is seen when you load in the package:

library(malf.templates)

Where is the Data located?

The malf_images is the general workhorse function for returning the data from the MALF templates. It takes one argument (group) to indicate which group from the competition to take the data (train or test). If both groups are passed, then all the data is returned. Group would generally only be specified if you wanted to do something relevant to the competition.

mi = malf_images()
names(mi)

The return is a list of filenames, broken into images (raw T1-weighted scans), labels (delineated labeled structures), brains (processed T1-weighted images, skull-stripped), masks (binary brain masks).

Derivative Functions

The malf_template_images, malf_template_labels, malf_template_brains, malf_template_masks functions all call malf_images and take the relevant element of the list for easier use.

MASS Derivatives

For backwards compatibility when the mass.templates package was integrated, the mass_images function was created. It simply returns the images and the brain masks. Generally, these are used for MASS (multi-atlas skull stripping), which is a bit more specific (but more widely-applicable) applicatio of label fusion:

mi = mass_images()
names(mi)
head(mass_template_images())
head(mass_template_masks())

The large difference is that mass_images (and the derivative functions) simply take the n_templates argument for the number of templates to return (from 1 to 35).

What to use them with?

Using the extrantsr::malf function, these functions can be used for MALF. The images/brain images are passed into the template.images argument and the structures/brain masks are passed into the template.structs argument.

Why have you not shown me a MALF example?

We have not included this code in the vignettes due to install times for extrantsr as a Suggests package.

neuroconductor/malf.templates documentation built on May 19, 2021, 5:24 a.m.