Predicting myelin in the brain by quantitative magnetic resonance imaging

Myelin is a structure that insulates neurons, allowing nerve impulses to travel along the axons at the appropriate speed for the maintenance of normal neuronal function. Little myelin is present at birth, however, it starts to develop rapidly during childhood, with myelination being essential for normal neurodevelopment. Diseases which cause myelin loss, such as multiple sclerosis (MS), require a way of diagnosing the extent of demyelination for monitoring and evaluation of novel remyelinating treatments. Thus, it is vital to be able to visualize myelin quantitatively. Current methods for the detection of myelin include the following quantitative MRI (qMRI) techniques: T1, T2 and T2* relaxometry; magnetization transfer, susceptibility-weighted imaging, and diffusion tensor imaging (DTI). Myelin is detected indirectly by measuring its effect on the local environment or its iron content. However, the correlation of each technique with histological myelin staining varies. The aim of this studentship is to establish which qMRI techniques best predict myelin, by comparing their measurements to corroborative histology and metal mapping data with machine learning methods. The qMRI measurements will be correlated to histological data and metal mapping data from the same brain tissue section to determine their accuracy and specificity. This will overcome difficulties with spatial registration of different image modalities by allowing point-to-point mapping of MRI and non-MRI data. The established techniques will then be tested in vivo to determine their accuracy, in both normal myelination model and in a model of myelin loss.