app-removeTractOutliers
This app uses mbaComputeFibersOutliers from the Matlab Brain Anatomy (MBA) script collection to perform a fiber cleaning operation. Specifically, it removes outlier streamlines on the basis of ether distance from fiber centroid core or distance from mean streamline length. Moreover, this app performs this operation iteratively across all fiber tracts mapped in the input White Matter Classification (WMC). For this reason, it is advised that this application not be applied to White Matter Classification (WMC) structures that do not correspond to "coherent white matter tracts", as the mathematical operations performed assume a "tract-like" morphology and coherence.
Authors
- Daniel Bullock (dnbulloc@iu.edu)
Contributors
- Soichi Hayashi (hayashis@iu.edu)
Project Director
- Franco Pestilli (franpest@indiana.edu)
Funding
References
Running the App
Inputs
This application reqiures a White Matter Classification (WMC) and corresponding tractography input. It is advised that this application not be applied to White Matter Classification (WMC) structures that do not correspond to "coherent white matter tracts", as the mathematical operations performed assume a "tract-like" morphology and coherence.
Configuration/Parameters
You can tweak the following parameters
- centroidSD
Cut streamlines which are this many standard deviations away from the tract centroid (geometric center). (default 4)
- lengthSD
Cut streamlines which are this many standard deviations away from the tract average streamline length (default 4)
- maxIter
The maximum number of iterations to compute tract averages and apply a cut. (try 5)
On Brainlife.io
Visit this page to run this app on the brainlife.io platform. A White Matter Classification (WMC) and corresponding tractography input.
Running Locally (on your machine)
Given that this is, in essence, a wrapper around the outlier removal function, it is recommended that users desiring to use the unedrlying code simply use the underlying functions (i.e. the wraper or direct function). This is as opposed to trying to use a dockerized version of this application. Alternatively, comperable methods are available via dipy.
Sample Datasets
Visit brainlife.io and explore the following data sets to find viable classification and tractography inputs:
HCP classificaiton: https://brainlife.io/project/5c3caea0a6747b0036dcbf9a/ HCP tractography: https://brainlife.io/project/5c3caea0a6747b0036dcbf9a/
Output
The output of this application is a White Matter Classification (WMC) structure which has been pruned in accordance with the set parameters.