Brief Review of Imaging and Hardware

• went over all of the magnetics concepts associated with MRI and hopefully how/why they are important to imaging, but didn't touch on all of the aspects of the MRI system, which you should be aware of, if you go into the MRI business.
  • how does the signal get from the rf coil(s) to the reconstruction computers
  • How do you interface to your rf coil/coil(s)
  • how do you integrate all of these bits and pieces together, what should go where
  • patient handling
  • how much power is required to provide the gradients that we like for imaging
  • no operator is going to see any of the stuff that we talked about. All they see is a GUI interface that lets them make images, so the best HW in the world is not so useful without a good interface.
• Use linearly varying magnetic fields to encode spatial information in the measured MR signal.
  

  $$\omega(\vec{x}) = \gamma \vec{G} \cdot \vec{x}$$ (10.1)

  
  

• Saw that in the presence of a gradient, signal and spin density can be related by the Fourier transform, if make variable substitution $k(t) = \gamma \!\!\!\!\!-\int_{0}^{t} dt'G(t')$.
  $$\begin{eqnarray*} s(\vec{k}) &=& \int d^{3}r \rho(\vec{r}) e^{-i 2 \pi \vec{k}\c... ..._{x}x + k_{y}y + k_{z}z)}\nonumber\\ &=& {\cal F}[\rho(x,y,z)] \end{eqnarray*}$$
  
  
  
  
  
  
  

  
  
  
  
  
  
  
  

• Say want to make 256x256 2D image, $T_T = N_{PE} T_{R}$ implies that $T_T \approx 4 min$. Reasonable. Use slice interleaving to collect 10-20 slices at the same time. Get images covering volume of interest in less than 5 minutes.

  
  
  
  
  
  
  
  

• This is OK for head that stays still, but not great for much of body, with lots of motion. If want to collect multiple contrasts, time can still pile up quickly.

• If want true 3D imaging will need to understand some new things.