The MR scanners have a limit to their resolution based on the strength of the magnetic field generated. The common 1.5 Tesla has a resolution of 1 mm or larger at best. This is fine for imaging bleeds and aneurisms but not even close for neurons or axons.

It takes a 5 or 6 Tesla scanner to start to see concussion damage finer than a bleed.

My point was that absent the symptoms of a bleed or other serious damage, the MRI is not much help except for the psychological value of ruling out such damage.

An fMRI can show areas of decreased brain function as compared to normal expected functions. There is some value to an fMRI study but only with skilled doctors setting up the protocol. fMRI is still mostly a research protocol.

There was a movement toward Diffusion Tensor Imaging and Diffusion Weighted Imaging. These techniques still need a 3 Tesla or stronger magnetic field to image the fine damage. I have not read anything lately about putting DTI or DWI into common use for concussion or such.

Some doctors use imaging, whether CT or MRI, as a liability limiting diagnostic. As greenfrog's doctor said, CT is more useful at the earlier stages, specifically, at 72 hours post injury when they will show any brain bleeds before they get absorbed. Earlier than 72 hours and the fine brain bleeds have often not become large enough to be readily recognized.

Some patients get an MRI/MRA. This is a dual MRI that by subtracting one image from the other, shows the vascular structure of the brain only. These will show the larger vascular structures but not the fine capillaries. I have a set of such images of my brain.

My neurologist suggested an MRI/MRA and SPECT (FDG PET scan) due to the severe brain dysfunctions he observed from the qEEG performed in his office. Both came back negative except for some minor out-pouching that appeared stable.