MRI multi-site reproducibility study
Despite careful calibration and quality assurance, MRI scanners from different vendors and at different sites produce differing images. This is a challenge for advanced methods aimed at imaging tissue microstructure. Individual standardised pipelines would be highly valuable in running multicentre MRI trials. Reproducibility is critical when translating the approaches into clinical application, and is the first step in allowing more hospitals and institutions to participate in studies of advanced MRI methods.
This project assesses the variability of advanced diffusion MRI studies of the brain in healthy volunteers performed at University College London, Cardiff University and University of Cambridge. The results of the work will enable a quantified understanding of how reproducible and reliable tissue microstructure images are when measured by different teams at different sites.
Quantitative analysis MRI phantom for longitudinal assessments
With the growth of imaging modalities based on quantitative analysis of MRI data, it is imperative that there also exists a mechanism to apply an uncertainty to any stated quantity. For the successful transition of these techniques into the clinical environment, a quantitative measure of inter-site scan data quality is necessary in order to compare results taken in different hospitals. A suitable test object must be developed that allows for the characterisation of fundamental MR parameters to be measured in a repeatable, traceable way.
This project will develop a novel quantitative MRI phantom, which will be used to perform a nationwide study of fundamental MRI capability through longitudinal multisite studies. This information will allow for an understanding of quantitative uncertainties throughout different imaging modalities and will lend confidence to drug trial studies, biomarker analysis and machine learning based algorithms.
MRI T1 relaxation time phantom development
Although phantom devices for calibration of MRI scanners can be purchased from a small number of providers, they suffer from a number of drawbacks including that the T1, relaxation time, values of the phantom materials are not guaranteed by the manufacturers, and that the geometry of the phantoms are typically only designed for head coils, so cannot fit within coils designed for imaging other parts of the body.
To resolve these and other issues, this project is developing magnetic resonance imaging phantoms suitable for specific anatomies. This will ensure that MRI scanners being used in clinical trials are equally and accurately calibrated and that they behave in a uniform manner throughout the lifetime of the trial. These phantoms are needed for multi-centre trials all over the world to ensure that the quality of the trials are of the highest standard.
Duchenne muscular dystrophy biomarker
Duchenne muscular dystrophy is an invariably fatal genetic, muscle-wasting condition affecting approximately 1 in 3,600 boys in the UK. The genetic basis of the disease is well-understood and a number of potential therapies have been developed, but the comparative rareness of the condition means that regulatory approval for new therapies is challenging to obtain because of the difficulties in powering a clinical trial – there are simply not enough patients to be able to draw robust conclusions given current effect sizes.
One way to address this is via more effective imaging. By partnering with researchers from University College London and Great Ormond Street Hospital, this project aims to demonstrate that fractional diffusion MRI imaging is a practical and feasible method for detecting microstructural tissue change in a clinical setting, thus supporting the translation of new treatments for Duchenne muscular dystrophy.
Tractography in Neurosurgery
Tractography is a technique for reconstructing white matter structures in the brain in vivo non-invasively from diffusion MRI data. The technique has developed significantly since it was first published in the 1990s but there is currently no way to attach confidence or uncertainty to white matter reconstructions. This is a serious barrier to routine clinical deployment as it is imperative that excising healthy tissue is avoided during brain surgery, since this causes permanent damage.
Partnering with imaging experts at University College London’s Institute of Child Health and a neurosurgery team at Great Ormond Street Hospital, this project will investigate how uncertainty in the imaging data affects tractography-based reconstructions of brain tissue, leading to new visualisation schemes and software for surgical planning and intra-operative monitoring that allow for robust estimation of uncertainty in tractography. Surgeons will be able to have more confidence in where healthy tissue structures are, and the extent of unhealthy tissue to be removed.
Assessment of the impact of a strong magnetic field on radiation response in MRI-guided radiotherapy
A recent development in radiotherapy combines magnetic resonance (MR) and linear accelerator (Linac) treatment units. This allows for improved imaging of the patient during radiotherapy, but the strong magnet is also likely to affect the distribution of the therapeutic dose at the tissue level, due to interactions between the magnetic field and the charged secondary electrons that mediate the effects of photon irradiation. In addition, at the nanoscale level, the presence of the magnetic field may have significant effects on the production and distribution of reactive radical species and cellular repair mechanisms.
This project will investigate how a strong magnetic field changes the cellular effects of radiation, in an advanced tissue engineering construct which recreates the in vivo environment. This will help to determine whether this new type of commercial radiotherapy unit, the MR-Linac, has radiobiological implications that can be safely discounted, compensated for or exploited, prior to clinical implementation.
