Exploiting the unique quantitative capabilities offered by simultaneous PET/MRI

Lead Research Organisation: University College London
Department Name: Medicine

Abstract

Both positron emission tomography (PET) and magnetic resonance imaging (MRI) are well-developed imaging methods that provide key diagnostic information in clinical studies. PET can detect very small quantities of radiolabelled tracers which illustrates ways in which the body functions and what is happening to cells and small molecules in the body. However, PET images have poor spatial resolution and can only be acquired in a relatively long time (several minutes). On the other hand, MRI provides high spatial and temporal resolution anatomical information unrivalled by any other imaging modalities, especially providing excellent soft-tissue contrast. It is widely recognized that imaging systems that incorporate two different imaging methods can add diagnostic value e.g. PET/CT. The UK's first simultaneous whole body PET/MRI system was recently installed at UCL/UCH providing a unique platform for cutting edge research. We aim to exploit the simultaneous capability of this unique instrument to improve the range of parameters available for clinical interpretation.
Our research programme will involve the following novel sub-projects.
a) We will take advantage of MRI anatomical and functional information to directly improve the accuracy and precision of PET reconstruction. This involves development of computer programs that can extract information on organ boundaries from MRI and use this information to control the PET reconstruction. In kinetic studies, where a time sequence of images is acquired, knowledge of the expected changes in time can also be used to improve image quality.
b) We will develop an efficient method for MRI motion estimation that can be applied for correction of simultaneously acquired PET data, in a practical setting. This involves the development of new methods of acquiring MRI data so that motion can be measured along with the standard methods that are required for clinical acquisition. The information on motion will be incorporated in the software used for the PET reconstruction.
c) We will develop joint solutions for simultaneously acquired PET and MRI kinetic data that has potential to benefit both modalities. A number of useful parameters can be extracted from dynamic studies performed either using PET or MRI, but there are usually quite large errors involved. No-one has previously used the information available from both techniques to improve the estimation of parameters. This is quite challenging but will permit the measurement of more specific parameters that may be more helpful in understanding disease and its treatment.
d) We will adopt an all-inclusive framework for PET reconstruction that encompasses the complementary MRI information: spatial information, motion estimation and kinetic models. This will attempt to utilize all the available information to improve the PET image quality and the reliability of quantitative parameters that can be made available to clinicians. We hope to extend this approach to a truly combined reconstruction of parameters using information from both modalities.

The proposed research necessarily involves a multidisciplinary team of researchers with complementary skills, who are well placed to direct the research to meet well defined goals. The developed techniques will be evaluated in patient studies performed on the new PET/MRI system as a basis for their future use in clinical research.

Planned Impact

Novelty of the proposed research:
We propose an ambitious project which directly targets a new area of diagnostic imaging as the motivation for further development of novel data acquisition, image reconstruction and image processing. We extend on the experience and strengths across several UCL departments in order to address some challenging problems. The appeal in our proposed all-inclusive framework for image reconstruction is that it provides a flexible approach to capture the unique properties of the combined PET/MRI system with potential to identify novel biomarkers that have immediate application.
Health and societal impact:
This project is centred on new technology only recently made available to clinical research communities and available for the first time in the UK. An important aspect of this technology is the reduced radiation exposure compared to PET/CT, particularly important in paediatric studies and longitudinal studies that involve multiple examinations. Exploiting the unique capabilities of this technology has therefore potential for significant impact on investigative medicine and clinical practice. The research seeks to identify and validate improved, more specific, biomarkers that permit optimally directed treatment; these can potentially also inform on disease progression and treatment response across a range of diseases and conditions. There is therefore potential for significant health impact and societal benefit.
Engagement with medical specialists is intrinsic to the project, with two imaging specialists included as co-investigators. The goal is to highlight utility of the developed algorithms in demonstration clinical studies and to provide pilot data that supports more extensive clinical evaluation. We have included costing for a scientist with specific responsibility for efficient software implementation, to ensure that translation of software to the clinical research platform is efficient, bridging a common gap in technological research. The focus of the proposed research aims to exploit unique capabilities of the recently released simultaneous PET/MRI system. The development of new acquisition strategies and analysis protocols will facilitate greater insight to the underlying molecular and biochemical processes using non-invasive in vivo procedures. Development of these specific techniques therefore has potential to lead to enhanced understanding of disease and its treatment. The ability to better direct and monitor treatment has important economical and quality of life implications.
Training and education:
In addition the proposed research will be undertaken in an environment that promotes cross-modality training and skills development in an area where there will be future employment opportunities. Optimal use of the combined modalities requires advanced knowledge of both PET and MRI, expertise which is currently quite rare. Engagement of young researchers in the early development of the dual modalities will provide a nucleus of expertise to aid future development. Appointing PhD students in strategically selected projects to further extend the project team (through independent funding) provides an avenue for further shaping the UK capability to support this new technology.
Imaging by its nature is very visual and we will use outputs from our research for public engagement. We will seek opportunity to exhibit our work at open days and exhibitions and to present our work to schools and the media, activities where we have previous experience. As this is the only facility of its kind in the UK, we foresee considerable press interest.
In summary the proposed project:
- involves development of novel acquisition and processing techniques
- is timely, with potential to influence future clinical utility of PET/MRI
- has potential for significant societal benefit with direct impact on health care
- provides a nucleus for multi-disciplinary training on multi-modal imaging

Publications

10 25 50
publication icon
Arridge S (2018) Variational Gaussian approximation for Poisson data in Inverse Problems

publication icon
Burgos N (2013) Attenuation correction synthesis for hybrid PET-MR scanners. in Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention

publication icon
Burgos N (2013) Attenuation correction synthesis for hybrid PET-MR scanners. in Medical image computing and computer-assisted intervention : MICCAI ... International Conference on Medical Image Computing and Computer-Assisted Intervention

publication icon
Burgos N (2015) Multi-contrast attenuation map synthesis for PET/MR scanners: assessment on FDG and Florbetapir PET tracers. in European journal of nuclear medicine and molecular imaging

 
Description We seek to exploit the simultaneous acquisition of PET and MRI studies. We are exploring several avenues and have already demonstrated potential in a number of areas:
- we have developed means of joint PET/MR reconstruction that improves on results of both modalities; the approach has been further extended as a promising alternative for incorporation of anatomical information to improve reconstruction; the approach is highly cited and is forming the basis for ongoing research.
- we are utilising MR data to improve the quantitative potential of tracer kinetic studies; this still needs to be demonstrated with clinical data. Further joint PET/MR modelling is being devloped with the aim to simplify the practical procedure by reducing acquisition time.
- we have developed techniques that obviate the need for invasive arterial blood sampling as part of kinetic analysis and aim to extend this to studies where there are blood-based metabolites. This approach is being applied in ongoing clinical research brain PET studies.
- we have developed methods to improve motion detection and correction during clinical PET/MR studies; this is to be translated for clinical application
- we have developed methods for automatically dealing with mismatch between emission and transmission studies that allow more patient-friendly acquisition; the continuation of this work is funded by GE Healthcare
In addition we have developed work that has immediate use in improving the quantitative analysis of PET/MRI studies:
- we have improved on the method of attenuation correction; continuation on this work is being funded by Siemens Healthcare
- we have implemented methods for partial volume correction which are being applied to clinical research studies
- we have implemented methods to facilitate the use of developed algorithms on real patient data - also inspired further work on joint reconstruction and registration for automated emission/transmission alignment for respiratory gated PET data.
Exploitation Route Findings are being published. Software being developed will be available open-source. On-line automated processing also made available to other sites. Also several subtopics are being continued under industrial support. Continues to inspire additional research directions.
Sectors Digital/Communication/Information Technologies (including Software),Healthcare

 
Description There have been several areas of impact for the research undertaken. - Improved attenuation correction methods for PET/MRI are being incorporated in clinical research studies including a major project in dementia and in evaluation of epilepsy; users also have access to a web-based application which synthesizes CT based on a supplied MRI; this has been accessed over 1000 times. - a comprehensive comparative review on attenuation correction in PET/MRI rates the approach highly in comparison to other methods - partial volume correction techniques are being used for quantitative analysis of dementia and epilepsy studies - a database has been developed that other sites can use to evaluate methods of partial volume correction . - motion correction techniques have been evaluated in clinical studies and the technique is being refined for routine clinical use - several researchers have sought advice on use of the developed approach for determining an image derived input function; methods have been adopted in several centres
First Year Of Impact 2016
Sector Digital/Communication/Information Technologies (including Software),Healthcare
Impact Types Societal

 
Description European doctoral training
Geographic Reach Asia 
Policy Influence Type Influenced training of practitioners or researchers
Impact Local expertise has contributed to training of multi-disciplinary personnel on use of PET/MR; demand for such training is increasing. Provision of advice to staff at other centres in the UK that have recently established PET/MR centres.
 
Description Short PET/MR courses
Geographic Reach Local/Municipal/Regional 
Policy Influence Type Influenced training of practitioners or researchers
Impact Local expertise has contributed to training of multi-disciplinary personnel on use of PET/MR; demand for such training is increasing.
 
Description UCL PET/MRI symposium
Geographic Reach National 
Policy Influence Type Influenced training of practitioners or researchers
 
Description Computational Science and Engineering: Software Flagship Project Call
Amount £523,639 (GBP)
Funding ID EP/P022200/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Academic/University
Country United Kingdom
Start 06/2017 
End 02/2020
 
Description EC FP7
Amount € 4,600,000 (EUR)
Organisation European Commission 
Sector Public
Country European Union (EU)
Start 03/2013 
End 02/2017
 
Description EPSRC CCP
Amount £250,000 (GBP)
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Academic/University
Country United Kingdom
Start 10/2015 
End 09/2020
 
Description EPSRC Software Infrastructure
Amount £357,507 (GBP)
Funding ID EP/K03829X/1 
Organisation Engineering and Physical Sciences Research Council (EPSRC) 
Sector Academic/University
Country United Kingdom
Start 10/2013 
End 09/2015
 
Description Industrial PhD funding
Amount £150,000 (GBP)
Organisation Siemens Healthcare 
Sector Private
Country Germany
Start 10/2016 
End 09/2020
 
Title Atlas based synthetic CT 
Description Development of an atlas based synthesis of a CT from MR data can be used for attenuation correction of PET data. 
Type Of Material Data analysis technique 
Year Produced 2014 
Provided To Others? Yes  
Impact The technique reduces the bias in quantification of brain PET/MR studies and is being used in clinical research studies of dementia and epilepsy. 
URL http://cmictig.cs.ucl.ac.uk/niftyweb/
 
Title Joint PET/MR reconstruction 
Description Algorithms have been developed by incorporating structural information in reconstruction via level sets; these provide mutual benefit to PET and under-sampled MR reconstruction. 
Type Of Material Data analysis technique 
Year Produced 2014 
Provided To Others? Yes  
Impact Open access is currently being implemented. 
 
Title PV correction evaluation 
Description Synthetic data based on real clinical studies provides a ground truth for evaluating partial volume correction algorithms. 
Type Of Material Database/Collection of data 
Year Produced 2014 
Provided To Others? Yes  
Impact Work is in progress to evaluate different PV correction techniques with a view to identifiying the most promising methods for clinical implementation. 
 
Description National network on PET/MRI image reconstruction 
Organisation University of Edinburgh
Country United Kingdom 
Sector Academic/University 
PI Contribution An EPSRC CCP has been established to develop a software platform for PET/MRI reconstruction. This is led from UCL.
Collaborator Contribution The partners actively contribute to the collaboration and regularly discuss direction for the software develoment.
Impact An initial software platform for reconstruction of PET or MRI data has been developed and is now available to the research community.
Start Year 2015
 
Description National network on PET/MRI image reconstruction 
Organisation University of Leeds
Country United Kingdom 
Sector Academic/University 
PI Contribution An EPSRC CCP has been established to develop a software platform for PET/MRI reconstruction. This is led from UCL.
Collaborator Contribution The partners actively contribute to the collaboration and regularly discuss direction for the software develoment.
Impact An initial software platform for reconstruction of PET or MRI data has been developed and is now available to the research community.
Start Year 2015
 
Description National network on PET/MRI image reconstruction 
Organisation University of Manchester
Country United Kingdom 
Sector Academic/University 
PI Contribution An EPSRC CCP has been established to develop a software platform for PET/MRI reconstruction. This is led from UCL.
Collaborator Contribution The partners actively contribute to the collaboration and regularly discuss direction for the software develoment.
Impact An initial software platform for reconstruction of PET or MRI data has been developed and is now available to the research community.
Start Year 2015
 
Description Siemens support for PET/MR lung attenuation correction 
Organisation Siemens Healthcare
Department Molecular Imaging
Country United Kingdom 
Sector Private 
PI Contribution A followup project has been established to investigate methods to obtain accurate PET attenuation correction based on MR images of the lung. Supervision is shared across scientists and clinicians at UCL/UCH.
Collaborator Contribution The partners have contributed half funds towards 3 PhD students. They also provide access to development software and interact regularly on the project direction.
Impact Too early to report on outcomes as started in October 2016
Start Year 2016
 
Description Siemens/UCL studentships 
Organisation Siemens Healthcare
Country Germany 
Sector Private 
PI Contribution Joint funding of four PhD students working on PET/MR: significant output from the student group; regular discussions on research outcomes. Discussions are underway to fund a further group of PhD students to continue on PET/MR research.
Collaborator Contribution Joint funding of four PhD students working on PET/MR: loan of off-line reconstruction software, technical advice on accessing real patient data
Impact Publications, presentations at international meetings, software (listed elsewhere).
Start Year 2013
 
Title PET/MRI software: attenuation correction 
Description The software synthesizes an atenuation map based on MRI data for use in attenuation correction of PET studies. This is made available for on-line access but relies on a database of clinical studies that are not distributed. The software is being used in the clinical research setting. No ongoing funding supports this. 
Type Diagnostic Tool - Imaging
Current Stage Of Development Early clinical assessment
Year Development Stage Completed 2014
Development Status Under active development/distribution
Impact This improves on the vendor supplied methods which result in quantitative bias. 
 
Title PET/MRI software: bloodless input function estimation 
Description The methodology is being applied in an internal clinical research study but has also been provided to other groups for their application in a clinical research setting. 
Type Diagnostic Tool - Imaging
Current Stage Of Development Refinement. Clinical
Year Development Stage Completed 2016
Development Status Under active development/distribution
Impact The proposed technique supports kinetic analysis without need for arterial blood sampling. 
 
Title PET/MRI software: partial volume correction 
Description A toolkit for processing data to correct for effects of limited resolution that result in quantitative bias; applicable to processing of PET/MRI brain studies (including separately acquired PET and MRI studies) 
Type Diagnostic Tool - Imaging
Current Stage Of Development Early clinical assessment
Year Development Stage Completed 2015
Development Status Under active development/distribution
Impact Makes the method easily accessible for further evalution in the clinical research setting. 
 
Title PET/MRI software: respiratory motion correction 
Description UCL Enterprise are funding software translation to facilitate clinical evaluation. 
Type Diagnostic Tool - Imaging
Current Stage Of Development Early clinical assessment
Year Development Stage Completed 2016
Development Status Under active development/distribution
Impact Preliminary data demonstrates improved lesion contrast and lesion detectability, reduction in artifacts. 
 
Title Atlas based AC map for PET/MR 
Description An application is provided to provide a synthetic CT to match supplied MR data for the purpose of attenuation correction of brain PET data. 
Type Of Technology Webtool/Application 
Year Produced 2014 
Impact This software is available for sites to evaluate the effectiveness of the software. This faciltates cross-comparison with other approaches. 
URL http://cmictig.cs.ucl.ac.uk/softweb/
 
Title Joint PET/MR reconstruction 
Description Joint image reconstruction for PET/MR data currently restricted to 2D test data. The algorithm utilises level sets and provides mutual benefit to both PET and undersampled MR. The open version is current under test. 
Type Of Technology Software 
Year Produced 2014 
Open Source License? Yes  
Impact There is considerable interest from the user community in this novel software; the open source software will allow groups to test the algorithms in their own lab which should build confidence in the developed methods. 
 
Title PET/MRI respiratory motion correction 
Description Software to enable the use of a predictive respiratory motion model based on MRI to be applied to PET data in combination with data-driven motion detection. 
Type Of Technology Software 
Year Produced 2016 
Impact This has already been demonstrated to improve image quality with increased lesion contrast and detectability and reduced artifacts in clinical studies. This is currently being translated for further evaluation. 
 
Title PV correction software 
Description Software for performing partial volume correction of emission tomography data using the RBV / iterative Yang correction methods. URL currently being tested internally. 
Type Of Technology Software 
Year Produced 2014 
Open Source License? Yes  
Impact In addition to this software an open-source library of test cases is being prepared for validation purposes (as part of an EC COST project). 
 
Title STIR (additions) 
Description Comprehensive reconstruction tool for emission tomography; additions relate to the extraction of data from the Siemens PET/MR system. 
Type Of Technology Software 
Year Produced 2014 
Open Source License? Yes  
Impact The STIR software package has a relatively large user base; the additions are mainly of interest to the small but expanding community who have PET/MR and who wish to experiment with alternative reconstruction algorithms. 
URL http://stir.sourceforge.net
 
Title non-invasive input function determination 
Description The softwrae enables the determination of activity concentration in blood without need for arterial sampling. 
Type Of Technology Software 
Year Produced 2016 
Impact Currently invasive arterial blood sampling is required if quantitative kinetic modeling is to be used. This limits practice to specialised research centres. The development may encourage wider use in clinical practice. 
 
Description UCH open days 
Form Of Engagement Activity Participation in an open day or visit at my research institution
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Public/other audiences
Results and Impact UCH organise an annual open day at which INM promotes activities including research projects.
Year(s) Of Engagement Activity 2014,2015
 
Description UCL PET/MRI Methodology Workshop 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact This 3 day workshop focused on development of analysis methods for PET/MRI with the following objectives:
• To present relevant current research at UCL and other centres
• To provide a forum for networking and future planning
• To encourage national collaboration on PET/MRI methods
We had several international of high standing. The workshop was very well received.
Year(s) Of Engagement Activity 2016
URL https://www.ucl.ac.uk/nuclear-medicine/upcomingevents/uclpetmrimethodologyworkshop
 
Description UCL open days 
Form Of Engagement Activity Participation in an open day or visit at my research institution
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Postgraduate students
Results and Impact UCL based open days organised for postgraduate students and for Centre for Medical Image Computing.
Year(s) Of Engagement Activity 2013,2014,2015