We are working on an animal model to see the effects of some diet in terms of functional and structural connectivity. In DTI data, I observed that images scanned at each gradient look significantly different from one another in terms of size, intensity, internal structure etc. I just wanted to ask from peers and current researchers that is it normal thing or something unusual happened during scanning?

If it is just the head motion artifact or eddy current artifact, please suggest the best tool and /or method to remove or minimize the artifact.

Two images are attached to see differences.