MRI physics: research and development
Our research and development work in MRI physics is concerned with the application and development of advanced MRI techniques. Some of our work involves developing new MRI techniques or applications. Other areas relate to transferring research techniques into the clinical domain.
MRI research areas
The assessment of cerebral autoregulation using MRI
This project uses advanced MRI techniques to assess cerebral autoregulation. Autoregulation is the mechanism by which blood flow to the brain is maintained at a constant rate despite variations in the pressure of blood supplying the vessels. These vessels achieve flow regulation by altering their resistance accordingly. This mechanism can become impaired.
The measurement and spatial localisation of the autoregulation mechanism across the brain is not currently possible, but could give us insight into a wide range of neurological disorders and conditions. This work is being carried out in collaboration with the neurological physics group within the department.
The picture on the right shows an initial parameter map of autoregulation across the brain.
The imaging of musculoskeletal characteristics using advanced MRI techniques
In this project, MRI is used to examine the structure and function of the intra-articular synovial folds (IASFs) in the cervical spine, in healthy volunteers. The IASFs are thought to be involved in pain associated with whiplash injuries, and little is known about their composition. Some quantification is being carried out to determine the shape and size of these folds, and we hope that the project will eventually look at how the IASFs changed with age and injury.
Other aspects of this project have involved looking at the morphology of the talus bone in the ankle and see if the ‘squatting facet’ can be determined from 3D reconstructions. The presence of this feature may provide further information about the function of the ankle joint complex.
Clinical validation of MRI techniques in temporal lobe epilepsy
Functional MRI (fMRI) is used to assess language function in temporal lobe epilepsy patients in this project. This is essential prior to surgery and is currently carried out using the WADA test, an invasive and time-consuming procedure. We want to develop a consistent and reliable procedure for testing language using fMRI that should be much easier to carry out than the WADA test.
The technique has initially been optimised on healthy volunteers and is now being used on patients where these is ambiguity in the lateralisation of language function. The results from this study can be validated against the WADA testing that the patient will undergo.
We plan to extend this work to investigate the use of fMRI to test memory function in the same group of patients. This will involve developing new paradigms and methods of image analysis.
The picture on the right is of images showing language function in a healthy volunteer.
Validation of a proprietary functional MRI analysis method
This project was the first time we explored the procedures and methods involved in functional MRI. We sought to compare the Siemens method of fMRI data analysis against the gold standard, statistical parametric mapping (SPM), and to establish which method best meets the clinical requirements for a centre such as Southampton.
We observed a few differences between the results using the two techniques, mainly that the extent of activation was usually over-estimated by the Siemens method, although the location of activation was fairly consistent between the two techniques.
The assessment of body fat volume and distribution in obese men
This project looked to quantify the volume and distribution of subcutaneous and visceral fat in a group of obese men using a set of non-contiguous T1 - weighted abdominal MRI images. These measurements were then related to other obesity related parameters to predict the presence of fatty liver (hepatic steatosis) in obese patients.
Quantification was carried out using a program called Mimics (Materialise LV, Leuven, Belgium). This involved creating ‘masks’ that isolated areas of visceral and subcutaneous fat. The masks allowed quantitative information to be extracted, which allowed percentages of visceral and subcutaneous fat volume to be calculated.
The picture on the right shows total, visceral fat and subcutaneous fat masks on a central slice of the abdomen.
The automated diagnosis of emphysema in COPD from high-resolution CT images of the lungs
In this project, high-resolution CT images were used to look for a quantitative measure of emphysema, based on the percentage of lung voxels that were below a certain threshold of Hounsfield units (HU). This measure was compared to other measures of lung function and quality of life. An automated analysis method was devised using IDL (Interactive Data Language, Research Systems Inc, Boulder, CO, USA).
Development of cardiac MRI techniques
This area of research and development has arisen due to the recent installation of a dedicated cardiac MRI scanner (Siemens Avanto 1.5T) at Southampton General Hospital. Our MR physics service will be supporting this new development, which will involve optimising pulse sequences and developing image-processing techniques. The unit will scan a large range of patients with both congenital and acquired heart conditions, and should provide an opportunity to develop a nationally leading cardiac MRI service.
Meet the team
Dr Angela Darekar leads MRI physics research at Southampton. She has many years experience in MR physics and also leads our MRI physics service.
Angela is supported by colleague Efstathios Varzakis.
Medical physics department
Southampton General Hospital