Motion Robust quantitative MRI of the brain at 7T

1) Brief description of the context of the research including potential impact

This project will be addressing the effects of motion on quantitative MRI (qMRI) of the brain at 7T. The Multi-Parameter Mapping (MPM) quantitative MRI protocol allows physical parameters reflective of underlying tissue microstructure to be quantified in a time efficient manner with high resolution and whole brain coverage. Motion-related artefacts are often prevalent in MRI scans, which can take many minutes at a time. It can therefore be uncomfortable for volunteers/patients to remain still for full sessions of multiple image acquisitions. With the increased inhomogeneity of magnetic fields utilised in the 7T compared to lower strengths, volunteer movement between scans can cause artefacts when these data are subsequently combined with signal models to quantify physical parameters. In collaboration with Siemens, our industrial partner for this project, Prof. Callaghan's group previously showed that motion is problematic for qMRI, not only because of encoding errors, but also because of position-dependent transmit and receive fields. Eliminating motion sensitivity from the MPM qMRI protocol would benefit clinical studies that seek to identify specific signatures of neurodegeneration, including ongoing studies in the Department of Imaging Neuroscience investigating epilepsy, Huntington's, Parkinson's and familial Alzheimer's Disease.

2) Aims and Objectives

The specific objectives are to:

Compare the accuracy of receiver coil sensitivity maps estimated either from fully sampled k-space areas integrated within the accelerated acquisitions or from short, low-resolution pre-scans; and to confirm that either dataset can be used to perform high-quality reconstruction and inter-scan motion correction in place of both to improve efficiency.

Conduct an investigation into the MPM-protocol in line with Balbastre et al, but with inter-scan motion correction performing either with the net relative sensitivities of the coils estimated between scans with the s-maps or using the net sensitivities estimated within the MORSE reconstruction. In this work we will assess the residual sensitivity to motion induced changes within the transmit and receive field of both approaches; and update the MPM protocol that is more robust and time efficient.

Alternative reconstruction scheme JLORAKS will be deployed to provide comparison to MORSE reconstruction.

Build on the qualitative assessment of MORSE and JLORAKS reconstruction of existing MPM datasets by focussing on a larger quantitative comparison of suitable multi-contrast schemes. I will determine the maximum acceleration factor that could be achieved without compromising image quality, including CAIPI sampling for optimal regular under-sampling. I will test various under-sampled protocols in-vivo in a group of healthy volunteers, with and without deliberate motion, and assess the image reproducibility of the MPM metrics.

Promising multi-contrast image reconstruction approaches will be further pursued and developed, now considering the potential for alternate under-sampling schemes where development of time-efficient image reconstruction would follow. A parallel stream of work exploiting the multi-contrast nature of the MPM protocol will focus on the covariance of the signal amplitude and common effective transverse relaxation rate of the constituent volumes.

3) Novelty of Research Methodology

This project will take multiple avenues of investigating ways to produce more robust and efficient protocols for scans; combining previous work from multiple sources as well as new untested methods.

4) Alignment to EPSRC's strategies and research areas

This project is heavily aligned with the EPSRCs goal by developing paths to shorten MRI scan time to achieve earlier and more effective diagnoses of both mental and physical ailments.

5) Any companies or collaborators involved

Siemens