Collaborative Computational Project in Positron Emission Tomography and Magnetic Resonance (CCP PET-MR)

The UK has been at the forefront of many innovations in both positron emission tomography (PET) and magnetic resonance (MR) imaging, making the UK one of the world leaders in medical imaging since its inception. The recent integration of PET and MR into a single simultaneous imaging system opens up exciting potential for synergistic imaging, and this timely proposal of a collaborative computational project aims to bring together the best of the UK’s PET and MR image reconstruction expertise to capitalise on this unique opportunity.

This new CCP PET/MR has been funded by the EPSRC from 1 April 2015 - 31 March 2020: with official website at:

The PI is Kris Thielemans (UCL) and the CoIs are Julian Matthews (UoManchester), Andrew Reader (KCL), Claudio Prieto Vasquez(KCL) and David Atkinson (UCL)

This collaborative computational project (CCP) in synergistic PET-MR image reconstruction has two primary aims:

1) PET-MR Networking

The UK has a number of research groups with high international standing, working on image reconstruction and analysis for either PET or MR. We aim to bring these groups together towards the goal of synergistic PET-MR image reconstruction and analysis by advancing understanding of PET-MR and the algorithms used for each modality. We will foster links with the wider international PET and MR communities to further achieve this goal.

2) Developing PET-MR Software Infrastructure

We aim to provide enabling infrastructure for research in PET-MR by development and promotion of a common software framework to tackle the specific challenges of PET-MR imaging. This will be achieved by standardisation of data formats, creating tools to export data in these data formats and development of a software platform for integrated PET-MR image reconstruction.

Mailing List

There is a JISCMAIL list at: easy to sign up by following this link.

There is an email for information: ccppetmr at


  • Initial kick-off meeting with the CoIs was held on the 1st of April at UCL
  • Meetings of Executive Committee to assign tasks and prepare Working Group meeting are ongoing
  • 27-30 April 2015 Physics & Software PET-MR training school in Leeds. Together with the EU COST action TD1007: "Bimodal PET-MRI molecular imaging technologies and applications for in vivo monitoring of disease and biological processes", we co-organised and co-sponsored the 2nd International Training School on PET-MRI Engineering. This 4 day school was attended by 33 trainees (9 from the UK).
  • Initial meeting with the Working Group was held on the 13th May 2015 11:30 to 15:00 at St Thomas' Hospital, London. We had 19 attendants from academia, representing all the UK groups working on (or interested in) PET-MR data processing and 5 attendants from industry (GE, Imanova, Philips, Siemens). Presentations, minutes etc are available from the main website (Working Group tab)


  • CCP SP at Daresbury Labs 24 June 2015

Future Events:

UK workshops:

  • Q4 2015: 1-day workshop jointly organised with CCPi: Synergistic Reconstruction of Multi-modality and/or Multi-spectral data
  • Q2 2016: 2-day workshop: Issues in PET-MR Reconstruction and Quantification (educative sessions, software training, research presentations)

Developer’s Meetings:

  • Tuesday 7th of July 2015 (1-5pm at UCL): software design of overarching framework
  • September 2015: Training: Software Carpentry (git, compiling, good programming practices)
  • November 2015: framework update and first release preparation

International activities:

  • Nov 2015: STIR User’s Meeting at IEEE Medical Imaging Conference
  • Nov 2015: Satellite Workshop at IEEE Medical Imaging Conference Open Source Software in Medical Imaging

Reading Materials

PET References:

MR References:

  • Jeffrey A. Fessler Model-Based Image Reconstruction for MRI 05484183.pdf
  • Image Reconstruction: An Overview for Clinicians Michael S. Hansen and Peter Kellman JOURNAL OF MAGNETIC RESONANCE IMAGING 41:573–585 (2015) Hansen_et_al-2014-Journal_of_Magnetic_Resonance_Imaging.pdf
  • Non-Cartesian Parallel Imaging Reconstruction Katherine L. Wright, Jesse I. Hamilton, Mark A. Griswold, Vikas Gulani and Nicole Seiberlich, PhD? JOURNAL OF MAGNETIC RESONANCE IMAGING 40:1022–1040 (2014) Wright_et_al-2014-Journal_of_Magnetic_Resonance_Imaging.pdf
  • MRI Temporal Acceleration Techniques Jeffrey Tsao and Sebastian Kozerke, JOURNAL OF MAGNETIC RESONANCE IMAGING 36:543–560 (2012)

Action List (from 1 April 2015)

Setup and access to web-site (potentially just IP address access to put some content on there), mailing lists etc.

The PET reconstruction engine will initially be based on STIR, Need to merge in a few local branches first (e.g. clean-up of GE contributed code) before open git access.

For MR, we will this year (2015-2016) target the Gadgetron.

For both, our intention is to create a user-friendly level "above" these such that at later stage we can add support for other packages. The super-level will need to have MATLAB support. It's still unclear if other target languages are desired. It could be that we create the "super-level" in C++, and then use SWIG to provide the MATLAB/Python/etc interface for instance.

Regarding SWIG. "mainline" SWIG doesn't have any MATLAB support yet but good progress is being made A fork from KT is on works for adding MATLAB support to STIR. Brief summary of current status of MATLAB support in SWIG:

- a lot of things work. It works for STIR, and STIR is relatively complicated.

- there are some memory allocation problems such that at exit of MATLAB, or when clearing STIR variables/functions in MATLAB in the wrong order, MATLAB can crash (but I haven't seen a crash during normal usage anymore since quite a while).

- the SWIG test-suite for MATLAB is incomplete but does throw up some errors which will need fixing. A lot of these are on support for "directors" (to be able to extend C++ from in MATLAB), which is a "nice-to-have" feature, but not essential for us at this point.

- SWIG MATLAB Documentation is minimal right now

- the SWIG maintainers rightly insist on the test-suite to work, and having TRAVIS support. The latter is tricky as TRAVIS doesn't have MATLAB (no license). There has been some discussion on getting minimal "run swig + compile" support for this on TRAVIS (i.e. without running the tests in MATLAB).

For conversations on SWIG, we will use the SWIG-devel mailing list, as we don't want to solve this on our own. A relatively old post is for instance,

We should use the STIR mailing lists for anything that applies to the general STIR audience. That way, we hopefully will get some help from the international community as well (and relief some of the pressure on me). STIR mailing list info is on the STIR website.


Kris Thielemans: k.thielemans at

Ron Fowler: ronald.fowler at

Martin Turner: martin.turner at

-- MartinTurner - 10 Feb 2015

-- KrisThielemans - 27 Jun 2015

Topic attachments
I Attachment Action Size Date Who Comment
pdfpdf 05484183.pdf manage 945.4 K 08 Apr 2015 - 08:42 MartinTurner  
pngpng CPP-logo2-DokChampa-cropped.png manage 13.4 K 19 May 2015 - 16:09 RonFowler  
pdfpdf Hansen_et_al-2014-Journal_of_Magnetic_Resonance_Imaging.pdf manage 609.4 K 08 Apr 2015 - 08:43 MartinTurner  
pdfpdf Wright_et_al-2014-Journal_of_Magnetic_Resonance_Imaging.pdf manage 1124.1 K 08 Apr 2015 - 08:45 MartinTurner  
pdfpdf corefunctions_c.pdf manage 4678.4 K 21 May 2015 - 19:47 MartinTurner  
jpgjpg p2015-05-13_onep40w.jpg manage 19.6 K 21 May 2015 - 18:53 MartinTurner  
Topic revision: r17 - 10 Aug 2015 - 13:49:34 - MartinTurner
This site is powered by the TWiki collaboration platformCopyright & by the contributing authors. All material on this collaboration platform is the property of the contributing authors.
Ideas, requests, problems regarding TWiki? Send feedback