Measuring Seeding Resolution Dependence of Diffusion Tensor Streamtube Visualization

We present a study to measure the relationship between seeding resolution involved in 3D diffusion tensor magnetic resonance imaging (DTI) tractography and the streamtube visualization of human brain DTI model. We investigate five scales of seeding resolution with the same brain dataset and measure the effect of the resolution difference on user performance on a set of tasks typical of those conducted by neurological experts. The study covers representative fiber bundles in the brain anatomy and the measurement metrics includes users’ task performance in terms of correctness, completion time, computer log, and participants’ comments collected from post-study interviews. Initial results show that lower resolutions bring about higher task correctness over all tasks consistently, and that the completion time fluctuates with changing resolutions and the pattern of change varies slightly from task to task.