![]() |
![]() |
![]() |
This page hosts various bits of code that we have used in the lab and links to bits of code that have been developed in collaboration with one or more members of the lab.
![]() |
Xtract
Xtract is a FSL tool for quick probabilistic tractography using pre-defined protocols. It allows you go take a library of standardized tractography protocols and quickly define tracts based on them in your own data.
Xtract comes with protocol libraries for the human and macaque, but the tool can be employed for other species. This means we can use it to standardize tractography protocols across species, which helps in formal comparisons, such as in the connectivity blueprint approach (Mars et al., 2018).
Xtract now comes with standardized human and macaque protocols, but more species are in development. It also serves as a type of spreadsheet for tractography, if you don't like our protocols you can just add your own and see how the results change. If you do, please let us know. We will continue to update protocols as our knowledge advances.
Details and validation Xtract are described in Warrington et al. (in preparation). The tool is found at https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/XTRACT.
Landscape-based clustering
Thomas Gladwin (University of Utrecht and Military Mental Health Research Centre, Netherlands Ministry of Defense) has developed a landscape-based cluster analysis for neuroimaging data, defining clusters intuitively by looking at their shape (second derivative of activation over space). An advantage of this approach is that this can be combined with whole-brain cluster-wise corrected tests without the need to define an arbitrary threshold for the initial definition of clusters.
The method is implemented as part of Thomas' hiro3 toolbox and described in a paper published in MethodsX.
Modified HCP pipelines
Lennart Verhagen adopted some of the minimal preprocessing pipelines released by the Human Connectome Project to make them compatible with the type of data we normally acquire. The code is placed on GitHub as the OxfordStructural fork of the HCP Pipelines.
It is solely developed for the processing of structural images. Primarily, in the fork the presence of T2w images is optional. Other additional features are: detection and masking of arteries, improved robust bias correction, correction of sinc interpolation errors, improved brain mask definition, and several small tweaks. Note that the fork is only tested on the FMRIB Jalapeno server in Oxford.