At the current resolution of diffusion-weighted (DW) magnetic resonance imaging (MRI), research groups agree that there are between one third to two thirds of imaging voxels in the human brain white matter that contain fiber crossing bundles. This thesis tackles the important problem of recovering crossing fiber bundles from DWMRI measurements. The main goal is to overcome the limitations of diffusion tensor imaging (DTI). It is well-known that imaging voxels where there are multiple fiber crossings produce a non-Gaussian DW signal. This is precisely where DTI is limited due to the intrinsic Gaussian assumption of the technique. Hence, this thesis is dedicated to the development of local reconstruction methods, segmentation and tractography algorithms able to infer multiple fiber crossing from DW-MRI data. To do so, high angular resolution diffusion imaging (HARDI) is used to measure DW images along several directions. Q-ball imaging (QBI) is a recent such HARDI technique that reconstructs the diffusion orientation distribution function (ODF), a spherical function that has its maxima aligned with the underlying fiber directions at every voxel. QBI and the diffusion ODF will play a central role in this thesis. There are many original contributions in this thesis. First, we propose a robust estimation of the HARDI signal using a closed-form regularization algorithm based on the spherical harmonics. Then, we estimate the apparent coefficient coefficient (ADC) to study HARDI anisotropy measures and to discriminate voxels with underlying isotropic, single fiber andmultiple fiber distributions. Next, we develop a linear, robust and analytical QBI solution using the spherical harmonic basis, which is used in a new statistical region-based active contour algorithmto segment important white matter fiber bundles. In addition, we develop a new spherical deconvolution sharpening method that transforms the diffusion q-ball ODF into a fiber ODF. Finally, we propose a new deterministic tractography algorithm and a new probabilistic tractography algorithm exploiting the full distribution of the fiber ODF. Overall, we show local reconstruction, segmentation and tracking results on complex fiber regions with known fiber crossing on simulated HARDI data, on a biological phantom and on multiple human brain datasets. Most current DTI based methods neglect these complex fibers, which might lead to wrong interpretations of the brain anatomy and functioning.