Hi when I opened a diffusion image of the brain in DSI studio I was able to see some 33 files.
Does a dMRI file captures 33 images at different instants of time?
Does a dMRI file captures 33 images at different instants of time?
It looks like this specific series acquired 33 volumes. The first are b=0, the subsequent 32 volumes each have b=1000 and different gradient directions. I think this article does a great job describing DTI. As it notes, a diffusion-weighted series must include at least 7 volumes: 6 diffusion-weighted images from 6 gradient directions (giving 6 values for Sk) plus one baseline b=0 image (giving S0).
The number of volumes in a diffusion dataset can vary. For example, the HCP study acquires 6 series, each with 6 b=0 images, and 90 directions sampled at three b-values (1000, 2000, 3000).
1 Like
Hey @Chris_Rorden thanks for the response. I have one question about streamline. I read that streamlines are a connected sequence of points. But streamlines is it actually predicted using a model or is the one derived from ground truths?
The streamline is predicted by a model. The simplest way to draw a streamline is to think of each voxel as a tensor, with the principle direction being V1. We can connect neighboring voxels together as a streamline when their V1 is close to co-linear and when the fractional anisotropy is large. This simple method will work for the center of large fiber bundles. More sophisticated methods help deal with locations where different fiber bundles cross each other.