Presentation on theme: "Chapter 9 Basic MRI I Mark D. Herbst, MD, PhD. Notice This lecture contained many drawings on the whiteboard, so get these from one of the other students."— Presentation transcript:
Notice This lecture contained many drawings on the whiteboard, so get these from one of the other students if you missed the class.
Spatial characteristics of the MR image The body is 3-dimensional The images we look at are 2-dimensional –Pixels have 2 dimensions –Voxels have 3 dimensions The third dimension is slice thickness Slicing and dicing the body for MR imaging is like going from whole potato to chips to fries to home fries.
K-space Signals from the patient are stored in the computer in a place called “k-space” and when it is filled, we use Fourier Transform to transform the information from k-space to “image space,” that is, an image.
Gradients We use gradients in 3 directions to get images –Slice selection gradient –Phase encoding gradient –Frequency encoding gradient –All these are orthogonal (at right angles) to each other
Slice select gradient Selective excitation –Apply a gradient in the slice selection (CC for axial slices, RL for sagittal slices, AP for coronal slices) Two determinants of slice thickness –Strength of gradient –Transmit bandwidth Remember the previous discussion about multislice imaging? We use the gradient to select the slice we want to excite with RF.
Slice Selection 3D volume acquisition –Uses two phase encoding gradient pulses, one in the usual phase encoding way, to be discussed below, and the other in the slice select direction. We need to collect a full set of echoes in a separate k-space for each slice, so we have to multiply the time for one slice by the number of slices to get total time. This is why we only use very fast GE methods to get 3D volume acquisitions.
Phase Encoding Gradient Applied after the excitation pulse (90 degree or alpha pulse) Changed to different level for each echo collected –If 256 x 256 image, need 256 echoes, therefore 256 different phase encoding pulses are applied, each at a different strength. –Acquisition of each of these echoes takes one TR if conventional spin echo, less if fast spin echo or EPI
Frequency Encoding Gradient Applied during the echo, can be seen on some pulse diagrams to be applied also during the FID. Changes the frequency of precession of the net mag vector in each row of voxels in the frequency encoding direction