Direct Measurements of Microstructure from MRI

The research will produce a new imaging paradigm called active imaging . Traditional imaging techniques are designed by physicists; medical or biological researchers use them if they provide useful contrast between different types of material or correlate with interesting effects. Recent trends in medical imaging are towards quantitative imaging techniques that combine biophysical models of tissue with traditional imaging techniques to provide more specific information relevant to particular applications. Active imaging extends this idea to exploit biophysical models more completely to design the imaging techniques themselves. More specifically, the technique uses optimization algorithms to search for combinations of images that provide the most information about the biophysical model and the best estimates of biologically relevant quantities.For example, Alzheimer's diseaseattacks and destroys brain cells. It leaves holes in brain tissue and deposits of unusual proteins. Brain tissue from Alzheimer's patients looks very different to normal tissue under a microscope, but the differences are not apparent on images from standard techniques like magnetic resonance imaging (MRI). Even techniques like diffusion-tensor MRI, which has acute sensitivity to tissue microstructure, show only moderate contrast. A broader class of technique, called diffusion MRI, measures the scattering of water molecules in tissue. The tissue microstructure controls the scatter pattern and so diffusion MRI provides information about the microstructure. Diffusion-tensor MRI provides only particular features of the scatter pattern that happen to be insensitive to the microstructural changes in Alzheimer's. However, we can tune the sensitivity of diffusion MRI in an almost infinite number of other ways. Active imaging will use a model of the microstructural changes in Alzheimer's to find the precise combination of diffusion MRI measurements that is most sensitive to those changes and discriminates them most successfully from normal tissue or other diseases.The project considers three diseases: Alzheimer's, multiple sclerosis and focal cortical dyplasia (a common cause of epilepsy). Each has characteristic abnormalities in brain tissue microstructure that current imaging techniques do not reveal reliably. The project will construct biophysical models of the abnormalities and use active imaging to devise diffusion MRI techniques that reveal them. The project will also use active imaging to tune diffusion MRI to reveal specific microstructural features of normal brain tissue, such as size and density of axons in white matter. No current technique can image these features in live subjects, but the information would provide fundamental new information about the structure and function of the brain. The active-imaging paradigm extends to almost any other imaging technique including other MRI techniques, X-ray or optical tomography or positron-emission tomography (PET). Although the project focusses on active imaging for diffusion MRI, it also aims to initiate follow-on projects to explore applications to other diseases (such as cancers) and other imaging techniques.