| Zugriffsnummer | 46472 |
| Dokumenttyp | Zeitschriftenartikel |
| Peer Review | mit Peer Review |
| Sprache | Englisch |
| Titel | Motion estimation and correction for simultaneous PET/MR using SIRF |
| Autor(in); Institution |
Brown, Richard; Institute of Nuclear Medicine, University College London, London, UK
Kolbitsch, Christoph; 8.1, Biomedizinische Magnetresonanz, PTB-Berlin
Delplancke, Claire; Department of Mathematical Sciences, University of Bath, UK
Papoutellis, Evangelos; Scientific Computing Department, STFC, UKRI, Rutherford Appleton Laboratory, Harwell Campus, Didcot, UK
Mayer, Johannes; 8.1, Biomedizinische Magnetresonanz, PTB-Berlin
Ovtchinnikov, Evgueni; Scientific Computing Department, STFC, UKRI, Rutherford Appleton Laboratory, Harwell Campus, Didcot, UK
Pasca, Edoardo; Scientific Computing Department, STFC, UKRI, Rutherford Appleton Laboratory, Harwell Campus, Didcot, UK
Neji, Radhouene; School of Biomedical Engineering and Imaging Sciences, King’s College London, London, UK
da Costa-Luis, Casper; School of Biomedical Engineering and Imaging Sciences, King’s College London, London, UK
Gillman, Ashley G.; Australian e-Health Research Centre, Commonwealth Scientific and Industrial Research Organisation, Townsville, AUSTRALIA
Ehrhardt, Matthias J.; Department of Mathematical Sciences, University of Bath, UK
McClelland, Jamie; Centre for Medical Image Computing, Radiotherapy Image Computing Group, Department of Medical Physics and Biomedical Engineering, University College London, UK
Eiben, Björn; Centre for Medical Image Computing, Radiotherapy Image Computing Group, Department of Medical Physics and Biomedical Engineering, University College London, UK
Thielemans, Kris; Institute of Nuclear Medicine, University College London, London, UK
|
| Quelle/Jahr | Philosophical Transactions of the Royal Society of London A: 379 (2021), 2204, 15 S. |
| ISSN | 1364-503X (PRINT) ; 1471-2962 (ONLINE) |
| DOI | |
| Verlag | London: The Royal Society |
| Zusammenfassung | 1. Introduction Multi-modality imaging has enabled a leap forward in medical imaging. The combination of information obtained from two or more physical processes can provide powerful information for diagnosis, disease staging and/or therapy monitoring. One relatively recent example of this is the combination of positron emission tomography (PET) and magnetic resonance (MR) systems into one integrated device. The first modality allows obtaining quantitative information on function and metabolism in vivo by measuring the distribution of molecules (“radiotracers”) labelled with positron-emitting radionuclide. The second modality measures magnetic moments of 1H to obtain anatomical and functional information, such as blood perfusion, blood flow velocities or diffusion. The combination of these two modalities has opened a range of new clinical applications and research opportunities, with current emphasis on brain and cardiovascular imaging. There is a growing number of these high-end devices (currently eight in the UK). It is being increasingly recognised and demonstrated that information from complementary modalities, and from multiple time points, can be successfully combined to deliver image quality benefits compared to conventional independent processing. One particular feature of PET/MR is that the acquisitions can be carried out truly simultaneously, giving the opportunity to use information from both modalities to characterise motion. Many methods exploit the simultaneity of MR and PET acquisitions, and the different properties of the two modalities for mutual benefit. MR generally provides superior structural contrast and better spatial and temporal resolution. Hence, advanced MR sequences combined with iterative reconstruction methods are often used to obtain images for motion estimation via registration. Such techniques allow time-resolved – or more often – gate-resolved (with a “gate” corresponding to a motion state) images to be obtained, with recent advances for joint cardiac and respiratory motion estimation and correction. However, the flexibility of MR means that often additional diagnostic acquisitions are required, precluding continuous MR motion acquisition, whereas PET data are available throughout the acquisition. Hence, other methods build motion models parametrised by surrogate signals derived from the PET or interlaced MR navigator data. These models can be estimated on part of the data such that other MR sequences can be used while PET data can still be collected, allowing correction for quasi-periodic motion due to respiration or head movement. From a researcher perspective, implementing these methods or developing new methods is very challenging. Although PET/MR manufacturers provide tools for data manipulation and image reconstruction, these tools may not have all the desired capabilities and are generally not flexible to customisation due to their proprietary nature. There is therefore strong interest in open source software (OSS) that can be used for some or all of the data processing. Examples of MR or nuclear medicine image reconstruction OSS include: • Gadgetron – MR • the Berkeley Advanced Reconstruction Toolbox (BART) – MR • the Software for Tomographic Image Reconstruction (STIR) – PET and SPECT • NiftyPET – PET • the Reconstruction Toolkit (RTK) – CBCT, CT and in the future SPECT • Customizable and Advanced Software for Tomographic Reconstruction (CASToR) – PET, SPECT and some CT support However, none of these packages can reconstruct both PET and MR data. We are therefore developing an OSS framework called the Synergistic Image Reconstruction Framework (SIRF). This development is led by the Collaborative Computational Platform on Synergistic Reconstruction for Biomedical Imaging CCP SyneRBI www.ccpsynerbi.ac.uk. SIRF was developed for synergistic PET/MR image reconstruction, aiming to exploit the rich cross-modality information during the reconstruction of both the PET and MR images. |
| Kostenfreier Zugang | Open Access Hybrid |
| Rechteinformation | CC BY 4.0 ; Creative Commons Attribution 4.0 License |
| Themenbereich der Metrologie | Metrologie in der Medizin |
Zitierung
Brown, R., Kolbitsch, C., Delplancke, C., Papoutellis, E., Mayer, J., Ovtchinnikov, E., Pasca, E., Neji, R., da Costa-Luis, C., Gillman, A. G., Ehrhardt, M. J., McClelland, J., Eiben, B., & Thielemans, K. (2021). Motion estimation and correction for simultaneous PET/MR using SIRF. Philosophical Transactions of the Royal Society of London A, 379(2204), 15 S. https://doi.org/10.1098/rsta.2020.0208