Diffusion tensor imaging in seven minutes: Determining trade-offs between spatial and directional resolution

摘要

Imaging protocols must obtain maximum information under tight time constraints, to minimize patient discomfort or attrition, and motion artifacts. As part of a pilot study optimizing DTI sequences for the Alzheimer's Disease Neuroimaging Initiative, we scanned 8 subjects with 3 DTI protocols of equal duration at two time-points (48 scans). If scan duration is fixed, collecting more diffusion-sensitized gradient directions can increase angular resolution at the expense of spatial. We compared 7-minute sequences with 3.0(48), 2.7(41), and 2.5(37) mm isotropic voxels (directions), to assess (1) SNR; (2) bias in estimating fiber anisotropy; (3) reproducibility over time; (4) intersubject variance--relevant for group comparisons. Statistical maps revealed that higher angular resolutions gave more reproducible estimates; FA depended on voxel size, with a steeper dependency in more heterogeneous regions. The intermediate resolution gave best SNR. 2mm DTI scans are common, but improved angular resolution may add temporal stability, and benefits for tractography.