MRI; Radiology Cafe notes Flashcards
(133 cards)
1
Q
Properties of superconducting magnet?
A
Accounts of B0.
6 toness
always on
2
Q
Which part of the superconducting maggnet makes a lot of heat
A
the constant current
3
Q
1 Tesla = how many Gaus
A
10 000
4
Q
The earths Gaus is what
A
0.5
5
Q
what are shiim coils?
A
lie just inside the outer main magnet. fine tune to make it as uniform as possible
6
Q
how many sets of gradient coils are there
A
3
7
Q
what creates the loud noise in MRi
A
rapid turning on and off of gradient coils
8
Q
Properties of the radiofrquency coils.
- orientation
A
They are at right angles to the Z/B0.
9
Q
Properties of the radiofrquency coils.
- they are tuned to what?
A
a particular frequency.
to also receive the MR signals.
10
Q
Properties of the radiofrquency coils.
- location
A
have to be as close to the body as possible
11
Q
What types of RF coil are there?
A
Transmit and receive - standard body coil - head coil Receive only - surface coil
Phased array coils
Transmit phased array coils
12
Q
What are phased array coils in MRI?
A
Multiple receive coils, receives signals individually but are then combine to imporve SNR.
13
Q
in which coils do the magnetic fields from tissues get recorded
A
RF coils
14
Q
what is precession
A
magnetic nuclei will rotate about the axis of the field
15
Q
What is equation for Precessional frequency
A
Larmor equation
F = gyromagnetic ratio x B0
16
Q
why do we flip the magnetisation to 90 degrees
A
the mangetic field strength can’t be measured in B0.
17
Q
How does the magnetisation get flipped 90 degrees?
A
usea rapidly oscillating magnetic field at 90 degrees to B0.
Called B1.
18
Q
What frequency does B1 needs to oscillate at? and why?
A
same frequency as the resonance frequency (1 Tesla - 42MHz) - ensures the most efficent transference of energy to nuclei.
19
Q
How long will nuclei remain in the transverse plane for?
A
for as long as the RF is applied, afterwards start to move back to longitudinal axis.
20
Q
How does the transverse magnetisation return to B0 direction?
A
spin lattice relaxation
spin spin relaxation
21
Q
An electric voltage is induced in the receiving coils. At what frequency will it be?
A
The larmor frequency. Nuclei precession.
22
Q
what is the value of T1
A
it is the time it takes for Mz to recover 63% of its value
23
Q
T1
what is it for water, fat and bone
A
Water - long T1 - fast moving molecules don’t absorb energy quickly so don’t cause much disruption. Water retains magnetisation.
Fat - short T1 - large molecules with low innate energy. Quickly absorb the energy and so quickly lose magnetisation.
Bone - fixed and rigid. least efficient ar removing energy from precessing nuclei. LONG
24
Q
what is FID - free induction decay?
A
the exponential rate of loss of coherance and net transverse magnetisation reduces to zero.
25
what is T2?
it is the time taken to lose 63% of its inital value.
26
what is T2 for water, fat and bone
The more relaxed the molecules (ie water) the more uniform the magnetic field. So retain coherance. Long T2
Fixed molecules disrupt the magnetic field more so lose coherane quicker. SHORT T2.
27
What is T2*
The fact that in the real world T2 is much faster than expected due to local and external magnetic field inhomogeneities.
28
Spin echo - what is the aim of spin echo?
to remove the T2* effects but maintain T2 contrast.
29
How does Spin echo worK/
90 degrees applied, time t do a 180 degree pulse.
2t do TE.
The reduction in signal form original 90 degress to 2t is the decay of the tissue due to T2 (removed the effect of T2*)
30
How many time is is TE repeated?
Hundreds of times in a sequence.
31
TE vs TR
TE is when the echo is received - it is 2t.
TR is the time until the next 90 degress pulse is put out.
32
What are the advantages of the spin echo sequence?
Truw T2 weighting
High SNR
Minimise susceptibility effects
33
Whare the disadvantages of a spin echo?
Long scan times
| More RF power needed than in GE.
34
in order to maximise T2 weighted images we can minimise T1 contrast at short or long TR. Why is along TR chosen>?
At a short TR the signals are too small to be of use.
35
in order to maximise T1 weighted images we can minimise T2 by using a short or long TE. Which do we use?
A long TE has no signal.
A short is therefore used.
36
For T1
| Typical T1 and T2 times
TR 300 - 600
| TE 10 - 30
37
For T2 weighted
| Tycpical T1 and T2 times
TR 2000
| TE 90 -140
38
slice selection is which axis?
Z
39
frequency encoding is which axis?
X
40
phase encoding is which axis?
Y
41
how does slice selection work?
- a gradient is applied in the Z (over the background field).
- Means each point along the Z will have a different field strength and therefore different Lamor.
- RF pulse to flip into the 90 degrees has to be the same as Lamor frequency.
- changing the RF pulse measn selecting a different slice along the Z axis.
42
what is resetting the gradient?
In relation to the Z axis gradient applied. Happens between selecting next slices to rephase the nuclei.
43
What is the RF bandwith?
the range of frequencies within the pulse
44
What is the RF pulse frequency?
is the frequency change that will move the slice selection up or down.
45
What is the gradient strength?
for the same bandwith a larger/steeper gradient will select a smaller slice.
46
what is read out gradient?
It is the x axis column
47
what are the properties of nuclei in the x axis? Why is this a problem?
different amplitudes
same frequency and phase.
The signal produced would create one big wave
48
How do we overcome the fact that all nuclei in the x axis have the same frequency and phase?
Apply another gradient in the x axis which creates a new frequency based on the location of the x direction
49
What is the problem of applying a gradient in the x direction? How do we correct this>
Causes the signals to become very small as they dephase.
To correct this we apply a dephase gradient first first which is followed by the read-out gradient
The gardient echo is the collection period that then receives the signal.
50
What are the artefacts of frequency enoding?
Aliasing and chemical shift
51
Aliasing - if the signal is not sampled regularly enoughed what happens
underestimate the signal
52
what is the Nyquist limit =
sampling frequency / 2
53
how does aliasing work in the frequency encoding?
under sampled will underestimate the frequency. As a result the lower frequency signals will be placed at the low end of the X axis.
54
How do we stop aliaising ?
Range of frequencies is limited to sampling using a band pass filter. (the receiver bandwith).
- use a larger FOV
- pre-saturation bands on areas outside FOV.
- anti-alisaing software
- switch phase and frequency encoding directions
- use surface coil (sensitive to FOV and improve SNR)
55
Frequency encoding edges of the FOV respond to what?
The limits of the receiver bandwidth.
56
What is chemical shift?
a difference in lamor frequency of fat and water is misregistered as a difference in location.
Only occurs in frequency encoding.
57
what factors effect chemical shift?
Higher magnetic field worsens as relative difference in frequencies is exagerated.
decreasing gradient strength will worsen - more frequencies in a single area.
narrower bandwith worsens -
58
What is phase encoding?
Apply a gradient across the y axis.
- the preivously all same frequencies and phase have different phases of that frequency.
- how much depends on the y axis direction location.
59
Why does phase encoding have multiple cycles?
Amplitude of phase encoding gradient is changed each cycle.
With multiple cycles a single point will have a different amplitude across each cycle.
- these varying amplitudes with each cycle create a wave.
Different location will have different frequency waves.
60
Describe how the location of the phase encoding is decided?
those at the far ends of the gradient will have maxiumum change in phases and so the highest frequencies.
61
How is K space ordered?
In each column is all the data from one frequency encoding step (X axis).
Each row is filled in by repeating phase encoding steps
Periphery is high frequency.
Centre is low
62
Low pass filter do
Only allows the lower frequencies.
Smooth image. No edges
63
High pass filter
Allows high frequency.
Sharp image no contrast.
64
Sequences
| - which two gradients are applied at the same time in order excite protons in a particular slice
- RF (along the Z)
| - Slice selection for the frequency along the RF gradient.
65
After a RF and slice selection are completed what is done next?
A phase encoding is done.
Need as many repeats as there are rows in the K space.
typically repeat 256.
66
what is the last step of the sequence?
Frequency encoding.
applied after each step of the phase encoding gradient.
During the gradient that the signal is read - hence readout gradient.
67
what is the equation for scan time?
TR x phase encoding steps (y axis resolution) x number of signal averages
68
in spin echo where does the 180 pulse come in?
Comes in at TE / 2.
Read out / frequency encoding signal is at TR.
The negative bit of the read out is done at same time as the phase encoding though (ie before the 180) - it is positive - which then after the 180 is now negative!
69
how are the MRI times made shorter?
Multi slice sequence
Within each TR run the cycles for all the slices at the same time.
Still repeat for phase encoding steps but as simulataneous across all slices reduces the time it takes.
70
what is multi echo dequence?
Within a single TR can have multiple 180 degree flips.
This will allow two TEs of the same row of K space.
Can create a PD and T2 image at the same time.
71
What is turbo spin echo sequence?
Do many 180 degree flips each with a different phase encoding.
72
What is echo train length?
number of echoes created within a turbo spin echo sequence.
73
What are the advantages of turbo spin echo?
Very fast
| can achieve two images of different contrast
74
What are the disadvantages of turbo spin echo>
heavily T2 weighted only
| differing contrasts
75
how is a graident echo different?
If a short TR is needed we need to cut down the scan time.
76
how does gradient echo work?
Flipped to an angle, returns quicker to M0 and can be flipped again.
77
what is the down side of using small flip angles
lower signal strength.
Mz is retained and recoves quickly.
In T1 weighted want a longer recovery time. So can be difficult to see this.
78
why can't you get T2 weighting with a gradient echo
you are not using the 180 degree pulse that removes the effects of T2*.
THerefore can get T2* but not remove the imhomogenities.
79
How do you start an inversion recovery?
with a 180 degree pulse
80
What is TI
wait until a time where it reaches 0 and put a 90 pulse here. TI. time to inversion
81
Diffusion - wighted imaging
| Diffusion coefficent means
the more a particle can move in a given amount of time
82
why is EPI used in diffusion weighted imaging
its very quick - stops motiona artefact
83
how does diffusion gradient work?
diffusion gradients are added either side of the 180 pulse.
one to dephase and the second to rephase.
- if water has moved it won't return a signal.
84
how many directions is the diffusion gradient applied in?
min 3 usualy 6-20
85
In diffusion weighted imaging what is the B value?
A more sensitive (higher B) value to molecular motion
86
Characeteristics of higher B value
more noise
more sensitive
less signal
87
how can you increase the b vlaue?
larger diffusion gradient (amplitude or duration)
| increased time between dephasing and rephasing
88
why is the Apparent Diffusion coefficinet necissary?
DWI is T2 weighted.
| A bright signal could be from T2 or restircted diffusion,
89
what is the difference between anisotropic and isotropic diffusion?
anisotropic is when diffuson is NOT equal in all driections - ie along nerve tracts.
90
what are the three types of diffusion tensor imaging?
fractional anisotropy
principal diffusion direction map
fibre tracking maps
91
Fractional Anisotropy map - whorks how?
Measures the asymmetry of diffusion.
0 - dark - symmetrical
1 - bright - asymettrical/ anisotropic.
92
Principal diffusion direction map displays how?
Colour based on direction
Brightness based on degree of anisotropy
93
fibre tracking map works how?
creates an image of the tracts
94
What is T2 shine through
DWI
- T2/ * weighting.
- Intrinsic high signal or restricted diffusion --> hard to tell.
ADC map removes T2 signal.
low signal on ADC map is truly restricting.
95
What is T2 dark-through
low T2 intrinsic signal might appear as high diffusion.
96
DWI Metal artefact
Because of T2* weighting very susceptible to metal artefact.
97
MR spectroscopy advantages
Identfiy metabolites in imaged tissue
98
MR spec disadvantages
Low resolution
Susceptible to heterogeneity of the magnetic field
Can only image a limited area
99
Name the commonly used metabolites
Myoinositol (glial)
Choline(cell membrane)
Creatine (energy metabolism)
GABA, glutamine, glutamate (intracellular neuronal transmitter)
100
name some other metabolites commonly used for MR Spec
N acetyl aspartate
Lactate
Lipids
Aminoacids
101
Benefits of single voxel MR Spec
High SNR
No specrtral contamination
short scan times
102
what are the disadvantages of MR Spec
Very small coverage
| low resolution
103
What is multi voxem chemical shift imaging?
spectrum is produced from multiple voxels.
104
What are the benefits of multi voxel chemical shif imaging
What are the disadvantages
larger total coverage
higher resolution
sepctral contamination from nearby voxels
low SNR
long imaging time
105
MRS uses the spin on what elements
Hydrogen
| also C 13 and P 31
106
What must be supressed in MR Sepc
Water - has high signal from proton count masks other weaker signals.
107
what ar e the time of flight effects?
Flow related enhancement
| High velocity signal loss or washout
108
What is flow related enhancement?
occurs in gradient echo imaging
result of magnetic saturation
gradient echo put a new pusle in before complete resolution of longitudinal magnetisation.
Multiple short TRs.
This reduces the overall transverse magnetisation that can happen.
Fresh blood arrives not supressed in this way. Has a large signal .
109
What is flow void?
occurs in spin echo
A alice gets a 90 and 180 RF pulse applied. If moving blood is only sibjected to one of these then it will cause a reduced signal.
110
What is the equation for flow void and how fast the blood needs to be going
If the tissue moves faster than TE/2 then all the initial 90 wont be subjected to the 180.
if slightly slwoer then some will generate a small signal.
111
how does spin phase effects occur (other angio tpye)
Using phase encoding gradients.
We know how long the gradient is applied, the magnitude and so we can work out the velocity of the proton moving along the gradient.
112
What are the types of magnetism?
Ferromagnetic
Superparamagentic
Paramagnetic
Diamagentic
113
Ferromagnetic
Strongest
- strong inside the magnetic
- maintains magnetic orietnation after removal from the magnetic
114
Superparamagnetic
Middle ground (some ferromagnetic properties)
115
Paramagentic
weakly attracted to magents
- weakly infleunces a magnetic field
causes quicker loss of magnetisation
116
examples of paramagnetic
oxygen
magnesium
gad
117
Diamagnetic
weakly repel the magnetic field
results in loss of signal
water
copper
nitrogen
barium sulphate
118
T1 paramagnetic contrast agents
Cuases enhancement of T1 and T2 relaxation.
Greater T1 and lower T2 signal.
T1 relaxation increases with Gad conentration.
After an optimum point - after this further increases reduce T1 as T2 more prominent .
- low signal in bladder as a result.
119
Hepatobiliary agents
Contain manganese
- paramagnetic
- T1 contrast agent
- intracellular
- not dynamic as taken up over 24 hours
120
T2 superparamagnetic contrast agents
speeds up T2 decay and reduces the signal, more reduction based on more concentration.
121
Example of T2 superparamagnetic contrast agent
Iron oxide
- removed by the reticuloendothelial system.
- injected one hour before images required
122
pixel area is =
FOV / matrix
123
voxel volume
pixel area x slice thickness
124
how can we avoid cross talk
image non contiguous slices ie 1 3 5
125
increaseing NEX increase signal by
the square root of eg double = 2
126
What are the two types of motion artefacts?
Patient motion
Flow artefacts
127
Patient motion causes what?
ghosting - low intensity copies of the oriignal image shifted in the phase encdoing direction.
128
Overcome patient motion
```
More signals
faster sequence
change phase and frequency direction
Gating
Breath hold
Navigator echo trigerring
```
129
What are the types of distortion artefacts?
Local field inhomogeneity
Non linearities in magnetic filed gradient
Boundaries between tissues
130
What is fat saturation artefact?
due to local inhomogentities, the lamor of fat might change - as a result it may not be fully suppressed n the fat saturation
131
Radiofrequency artefacts types
Improper coil selection - artefaacts from areas outside the FOV
spurious RF signals - poorly shielded room. Band of noise inthe phase encoding direction
132
Data collection artefacts
Inadequate FOV - phase wrap around
Spurious data - herringbone artefact
Edge representation - Gibbs or edge ringing artefact.
133
Sequence specific artefacts types
EPI - rapid imaging, distortion around inhomogeneities
FSE - long Echo train length, T2 relaxation at the end. Blurring in phase encoding direction
SSFP - local inhomogenities, banding pattern on the image with signal void and stimulated echoes that interfere
