• Iterative and non-iterative algorithms for tomographic imaging (PET, SPECT, X-ray CT).
• Methods for magnetic resonance (MR) image reconstruction, including compensation for off-resonance effects (field inhomogeneity, susceptibility, etc.).
• Methods for MR RF pulse design, including MRI spectral-spatial pulse design for phase-precompensatory slice selection.
• Iterative image restoration tools.
• Methods for B-spline based image registration with regularization to encourage the deformation to be invertible (diffeomorphic).
• A NUFFT Matlab toolbox that performs fast and accurate nonuniform FFT computations.

Downloading

• Via this link: http://web.eecs.umich.edu/~fessler/irt/irt you can access the individual files and directories.
  However, if you access them one-by-one, you will need to download many files to get much of anything to work.
• Instead, you should just download the entire package file via this link: http://web.eecs.umich.edu/~fessler/irt/fessler.tgz.
• There is now a github version but it has the code only, no data or mex files, so it is still a work in progress.

Installing

• After downloading, move the tar file to a clean subdirectory before extracting its contents since it installs into the current directory.
• Start by reading the readme file.
• Then try some of the examples in the examples/ directory.

License

• You may freely use and distribute this software as long as you retain the author's name (myself and/or my students) with the software.
• It would also be courteous for you to cite the toolbox and any related publications in any papers that present results based on this software. A typical citation is like: Jeffrey A Fessler, "Michigan Image Reconstruction Toolbox," available at [url], downloaded [date].
• UM and the authors make all the usual disclaimers about liability etc.
• If you make changes to any files, then please change the file name before redistributing to avoid confusion (like the GNU software license). Better yet, email me the changes and I'll consider incorporating them into the toolbox.

Notes

• If you would like to be on email list to receive announcements about updates then please email "fessler AT umich DOT edu". These will be VERY infrequent emails and I will keep your email address private. You can unsubscribe at any time by sending another email to me.
• Although Matlab is an expensive commercial software package, many parts of the toolbox will run in open-source alternatives to Matlab like GNU Octave and freemat. I hope someday it will all work with Octave.
• Code for reproducing results in publications is here: http://web.eecs.umich.edu/~fessler/irt/reproduce.
• We welcome feedback, but do remember that this is open source software so you should study the code before sending bug reports.

External links

• For circular (axial) cone-beam CT reconstruction, see Reconstruction Toolkit (RTK)
• For 2D and 3D cone-beam CT see the ASTRA Toolbox. from Univ. of Antwerp.
• For CUDA and OpenCL extensions see this paper by Miriam Leeser et al. and this source code.
• Tomographic Iterative GPU-based Reconstruction (TIGRE) toolbox has SART/FDK/TV etc. on with GPU versions
• The NUFFT code has been
  • incorporated into Mike Lustig's SPIRiT package for MR reconstruction,
  • and into Florian Knoll's MR reconstruction package.
  • and was compared to the new FINUFFT library.
• The learnCT JAVA tutorial may be of interest too.
• The "gadgetron" site provides open-source reconstruction code for MRI.
• The arrayShow tool is a Matlab image viewer that has been designed for the evaluation of multidimensional complex images.
• Cone-beam CT (CBCT) code by Kyungsang Kim that uses GPU
• Operator Discretization Library (ODL is a python library for inverse problems from Ozan Öktem at KTH