I recently wrote a paper on MRIs and their impact on the medical field.  You can read it here: MRI Paper

     This was before I learned about MRI image distortions and basically everything inconvenient about MRIs, so keep in mind that my views on this certain imaging device has changed since starting this project and the MRI isn't as great as I make it out to be in this paper (I mean the MRI is really useful, but there are many drawbacks).  The paper does provide a history of neurosurgery and some neat new derivatives of the MRI.  

     The MRI uses a magnetic field to create it's images.  No matter how good you are at building a magnet, the magnetic field will always be inhomogeneous, and because of this, the image can be distorted.  For example, on a CT scan, a circle being scanned will look like a circle on the image and distances are accurately scaled up and scaled down.  In an MRI, a circle being scanned might have a small indent and distances are not always accurately scaled up or down.

     Therefore, using MRI scans can pose a problem in stereotactic neurosurgery where the target cannot be missed by more than 1-2mm.  However, CT scans cannot show the anatomical soft tissue structures of the brain, only the MRI can.  I know.  It's a constant struggle.