MRI III and IV Flashcards
What is the formula for spin echo contrast?
S is prop to (rhoH)(1-exp(TR/T1))(exp-(TE/T2))
What are typical T1 values for fat, CSF, White, and Grey Matter?
fat: 260ms
CSF: 2400
White: 780
Grey: 900
What are typical TR and TE values used for T1 weighting?
shorter TR (400-600ms) short TE (3-10ms) - should be close to T1
Why would we use a T1 weighted image?
T1 generally gives more anatomical/Structural info
What is Proton Density Weighting? What do we use for Tr, TE?
We wish to minimize contributions from T1 and T2 so use long TR and short TE.
Compared to T1 image, white and grey matter are inverted. So the sulcus in the brain folds will be lighter than other matter
What are typical T2 values for fat, CSF, White, and Grey Matter?
fat: 80ms
CSF: 160
White: 90
Grey: 100
What values should we choose for TR and TE for T2 weighting?
Long TR (2000-4000ms) Larger TE (80-120ms)
Fill in the table of parameter choices for different weightings:
param………….T1………………..proton……………..T2
TR
TE
param………….T1………………….proton………………….T2
TR…………..400-600……….2000-4000……..2000-4000
TE……………….5-30…………………5-30…………………6-150
In which image (T1 weighted, T2 weighted, proton density weighted) is CSF the brightest? Grey matter? White matter? Fat?
CSF - T2, proton, T1
Grey - Proton, T1, T2
White - Proton, T1, T2
Fat - similar (we never want to image fat)
What is inversion recovery spin echo? Why would we use it?
180……TI……90readout……TE/2……180……TE/2
Since the first pulse is 180, the Mzs all become -Mz, and we can choose the 90readout pulse when one of the recoveries is at zero, getting NO signal from that! Fat supression!
What does STIR stand for? What can we see?
Short Tau Inversion Recovery. Fat.
What does FLAIR stand for?
Fluid Attenuation Inversion Recovery. I think it allows us to see fluids going places? Tumors?
What does slew rate mean? What is a typical slew rate?
Slew rate is how fast the coil can reach its max strength.
Typically 5-250 mT/m/ms
How are SNR and Bandwidth related?
SNR is proportional to (BW)^-1/2
What is “RePhasing”?
The equal and opposite area applied by the “rephasing gradient” during the slice selection pulse
How do we localize in the x, y, z directions? In what order do we do this?
z: slice selection by only exciting one frequency
y: phase encoding gradient
x: frequency encoding
z, y, x
How does phase encoding work?
We apply different phase shifts, very slightly changing the speed of rotation in different y locations. Vary this to get different lines. Strength of the pulse indicated phase shift.
You are always looking for a “special” amount of phase shift so apply different PEG strengths and this “special” amount will be a different y locations.
Do we get more info from looking at lower or higher frequencies?
More comes from lower (in the centre of the diagram)
How would we calculate image acquisition time?
t = (TR)*(#PEG steps)*(Nexcitations) t = (600ms)*(192)*(2)
= 3.74 mins per slice!!
What is a common pixel size for MRI?
In-plane resolution: 0.5 - 1.0 mm
Slice thickness: 5 - 10 mm
Why do we have such a large slice thickness for MRI?
signal is very small so need large slices for decent SNR
Would increasing B0 allow us to increase spatial res?
Yes! higher B0 means more signal so we can take smaller slices
What are some negatives of increasing B0?
more RF absorption - tissue heating
T1 gets too long
artifact production
Would the following increase or Decrease SNR?
Larger flip angle LongerTR, shorter TE Smaller pixel size Thicker Slices Larger FOV More averages
Larger flip angle - increase LongerTR, shorter TE - increase Smaller pixel size - decrease Thicker Slices - increase Larger FOV - increase More averages - increase
What factors is SNR proportional to?
intrinsic signal intensity from pulse sequence voxel size (# of repeats in same voxel)^1/2 (BW)^-1/2 f(quality factor of coil) f(magnetic field strength) f(slice gap) f(reconstruction filter)
Signal average is proportional to what?
proportional to square root of averages
if we were to double the number of excitations, we would double the acquisition time, and get a gain of how much?
40%
Why do we care about k-space symmetry
we use k-space symmetry to fill in the gaps thus reducing acquisition time
If we narrow our bandwidth, do we need a steeper or less steep gradient?
we dont need as steep of gradient because we are more discriminatory between adjacent slice frequencies. But, we would need a longer sampling time.
What happens when we put gradient coils closer to the patient? Further away?
Closer we get more patient loading (ie patient-induced eddy currents, meaning the coil can’t read as well).
Further, we get distortion artifacts
How much does SNR increase when we increase B from 0.5T to 1.5T?
SNR increases by 3-5 times!
Describe cross-excitation
Because the bandwidth is not perfectly rectangular, so some of the signal comes from adjacent slices. If we go slice by slice, this can lead to proton saturation. To avoid this, we “skip” slices and come back to them after.
Name 3 main types of artifacts
Machine dependent
Patient Dependent
Signal Processing Dependent
Discuss Machine-Dependent artifacts
Magnetic Field Inhomogeneities (Global and Focal)
Global and be fixed with shims and sheilding
Focal (caused by makeup, metal implants etc). This is when you are actually exciting protons in a different location than you think you are. Leads to signal voids, displacement, geometric distortions.
What are susceptibility artifacts? Where are they?
susceptibility is ratio of internal to external marnetic fields
they occur at tissue interfaces, especially near metal
some pulse sequences are more susceptible to this than others
What kind of distortion would a Gradient Field Artifact cause?
a stretch (oblong brain)
RF artifacts cause cross excitation. How can we get rid of them?
Interleaving
Add a gap
What causes “ghost images” and along which direction do we normally see them?
caused by motion artifacts. Often along FEG (x) direction.
What is a chemical shift artifact?
Are these worse for smaller or large B fields?
Are they worse for stronger or weaker gradient fields?
It occurs when there is a difference in precessional frequencies of water and fat. this happens because the rest of a fat molecule shields the hydrogen and produces a change in the magnetic field. Causes a shift in anatomy, sometimes signal voids.
Larger B0 makes these worse.
Weaker gradients make them worse because of the smaller bandwidth.
