UCL Institute of Neurology


Magnetic resonance physics

Professor Paul Tofts

The MRI Physics Group supports existing novel MR methods, and develops new ones. These are principally with application to MS, although there is close collaboration with other groups, particularly other neurophysicists based at the Epilepsy MRI Unit at the Chalfont Centre for Epilepsy.

This Group’s work represents the largest contribution to the recent book 'Quantitative MRI of the Brain', which won the prize for radiology in the 2004 British Medical Association book competition. It is hoped that with the planned new MRI facilities at Queen Square, the methods developed by the Group can be used in a whole range of neurological diseases to show the effects of disease and its response to treatment.

Quantitative methods form a large part of the new methodology. A large set of tissue parameters can now be measured in the brain. DTI is particularly exciting; this detects subtle degradation of tissue through the reduction in biological barriers to water self-diffusion. Tracking the direction of greatest diffusivity enables nerve fibre tracts to be mapped (‘tractography’), and has great potential for studying functional deficits. fMRI, also an increasingly important part of our ability to study the effects of disease, compliments tractography. Quantitative magnetisation transfer measurements enable restricted protons in myelin and other macromolecules to be characterised, even though they are invisible to conventional MRI. This technique is being formally optimised in terms of its accuracy and reproducibility. Inter-scanner effects are being characterised, to enable the techniques to travel easily between MRI centres. Other quantitative techniques include spectroscopy (which measures concentrations of metabolites such as nacetyl aspartate, a neuronal marker), and blood perfusion (flow), measured noninvasively with arterial spin labelling, which is altered in MS.

Novel analysis techniques are also productive, and are important to make best use of the large multi-parametric datasets that are now available. Volume and whole-brain histogram measures are used. Principal Components Analysis and Linear Discriminant Analysis for histograms can extract optimum features without bias. Long-term stability of measures, to enable long-term clinical studies to be undertaken, is an important and challenging goal. Through our recent MRI scanner upgrade, serial measurements have been largely stable. Several new techniques are planned. Grey matter imaging may show MS lesions which contribute to functional deficit, yet go undetected in conventional MRI. In the optic nerve, diffusion and magnetisation transfer measurements should give increased insight into optic neuritis, which often precedes MS.

For more information on related groups see our page on Neurophysics at the ION.