This project focuses on developing a technique for quantifying and characterizing myelin in the central nervous system, and using it to investigate the pathophysiology of multiple sclerosis and other demyelinating diseases. One of the most effective MR biomarkers for myelin is the water trapped between myelin sheaths. The signal emitted from this water compartment can be distinguished through its fast T2 decay time governed by magnetic dipole-dipole interactions between the mobile water protons and the bound protons in the myelin macromolecules. One approach to probe sub-voxel content, such as myelin, is multi-component T2 (mcT2) analysis - a deconvolution process that converts a T2 signal into a distribution of T2 values (T2 spectra).  In the white matter these spectra are useful for estimating the myelin water fraction (MWF) by calculating the relative area of the short T2 values (0…40 ms), providing a reliable proxy of myelin content. In our approach (Data-Driven) statistical analysis is applied to the entire WM as a preprocessing step in order to identify a set of global mcT2 features which are then used as basis-functions for voxel-wise mcT2 fitting.