NifTK is the name of our platform, combining NiftyReg, NiftySim, NiftyRec and NiftySeg via the viewer NiftyView. The aim is to take the tools that TIG/CMIC develops, and apply them in a clinical context.
NiftyView is a cross platform graphical user interface, providing an entry point to the above packages, and in addition, many of the usual features you would find in any medical image viewer. NiftyView provides a working environment for our clinical collaborators, and research scientists. Contributors include Matt Clarkson, Gergely Zombori, Miklos Espak, John Hipwell, Steve Thompson and Sébastien Ourselin.
NiftyWeb is a web service tool that provides an entry point to our cutting edge algorithms. NiftyWeb provides a friendly interface and allows to any clinician or research scientists to test or use our algorithms with their optimal configurations. The website currently runs on a distributed network, where different nodes are responsible to specific algorithms. Contributors include Ferran Prados, M. Jorge Cardoso and Sébastien Ourselin.
Test it here.
NiftySeg is an open-source software package for image segmentation, bias field correction and cortical thickness estimation. Image segmentation is based on an Expectation-Maximisation algorithm with a Markov random field for local consistency and an improved sulci delineation enhancement. Cortical thickness is measured using a Laplacian based metric within a Lagrangian-Eulerian framework. Contributors include Jorge Cardoso, Marc Modat, Matt Clarkson and Sebastien Ourselin.
NiftyReg contains programs to perform rigid, affine and non-linear registration of medical images. Two versions of the algorithms are included, a CPU- and a GPU-based implementation. Further information about the method can be found here. Contributors include Marc Modat, Pankaj Daga, David Cash and Sébastien Ourselin.
NiftyRec is a software package for fully 3D Stochastic Emission Tomographic Reconstruction. The package includes a Matlab Toolbox with inline documentation and a library for C/C++ development. Computationally intensive tasks have a GPU implementation based on CUDA. Currently MLEM, OSEM, and OSL-MAPEM reconstruction algorithms are supported. Contributors include Stefano Pedemonte, and Sebastien Ourselin.
NiftySim is a light-weight, cross-platform finite-element modelling package. NiftySim contains nVidia CUDA and C++ implementations of the Total Lagrangian Explicit Dynamics (TLED) algorithm and several non-linear constitutive models making it suitable for the fast and accurate simulation particularly of soft-tissue biomechanics. Numerous extensions to the basic TLED have been made as well, such as the addition of contact models, and membrane and shell elements. Contributors include Zeike A. Taylor, Stian F. Johnsen, Matthew J. Clarkson, John Hipwell, Johannes Totz, Marc Modat, Bjoern Eiben, Lianghao Han, Yipeng Hu, Thomy Mertzanidou, David J. Hawkes, and Sebastien Ourselin