High-resolution diffusion tensor imaging of the human pons with a reduced field-of-view, multishot, variable-density, spiral acquisition at 3 T.

Diffusion tensor imaging of localized anatomic regions, such as brainstem, cervical spinal cord, and optic nerve, is challenging because of the existence of significant susceptibility differences, severe physiologic motion in the surrounding tissues, and the need for high spatial resolution to resolve the underlying complex neuroarchitecture. The aim of the methodology presented here is to achieve high-resolution diffusion tensor imaging in localized regions of the central nervous system that is motion insensitive and immune to susceptibility while acquiring a set of two-dimensional images with more than six diffusion encoding directions within a reasonable total scan time. We accomplish this aim by implementing self-navigated, multishot, variable-density, spiral encoding with outer volume suppression. We establish scan protocols for achieving equal signal-to-noise ratio at 1.2 mm and 0.8 mm in-plane resolution for reduced field-of-view diffusion tensor imaging of the brainstem. In vivo application of the technique on the human pons of three subjects shows a clear delineation of the multiple local neural tracts. By comparing scans acquired with varying in-plane resolution but with constant signal-to-noise ratio, we demonstrate that increasing the resolution and reducing the partial volume effect result in higher fractional anisotropy values for the corticospinal tracts.