Structural neuroimaging Flashcards

1
Q

What happens if we use a stronger mag field?

A

When stronger mag field, resolution is higher but are more artefacts, you get more data (more useful info but also more noise)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
2
Q

What is there in the scanning room?

A

Magnet (static mag field)
Gradient coil
RF coil

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
3
Q

What are the effects of mag fields on the human body?

A

• Less obvious effect on human body: influence on nuclei
nuclei –> nuclear magnetic resonance imaging (NMR)
• ≠ Radioactivity
• Nausea (only at 7T)
• No consequent short
short-or long long-term illness

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
4
Q

What are the effects of mag fields?

A

Atoms act like tiny magnets, very sensitive to magnetic field, when go into scanner, all p+ exp the static mag field and align with direction of field and spin at Larmor f, f correlates with strength of field
• Most relevant element for brain imaging is hydrogen (1H)
• Protons align with direction of the magnetic field –> nuclear magnetic resonance imaging
• Protons spin at the Larmor frequency that depends linearly upon strength of magnetic field
• If additional magnetic field oscillates at the Larmor frequency, nuclei absorb energy from the field –> nuclear magnetic resonance imaging
Start with no field, all p+ are aligned randomly, when go in scanner, all p+ align in direction of field and spin at Larmor f, then apply oscillating field (RF pulse), makes p+ spin in same phase => get more É, then turn off RF pulse and p+ go back to original state (precession); get out of phase and realign with direction of static field and generates radio f signal, energy from 1p+ is small but adds up with all p+, the induced current is the MRI signal that we measure
• When RF pulse is no longer applied:
 De-phasing of atoms
 Re-aligning to static magnetic field (flip back)
 emits energy = small signal in radio frequency range
• Small signals over all the re-aligning nuclei integrate
• The less de-phasing happened, the stronger this signal is

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
5
Q

What is a gradient mag field?

A

Remember: Larmor frequency depends upon field strength
• Static magnetic field varied across space
• Nuclei in different locations have a different Larmor frequency –> RF pulse only affects nuclei with matching Larmor frequency
• Three orthogonal gradients of field strength applied on top of static magnetic field
• Slice selection gradient: applied at time of RF pulse
• Phase encoding gradient: use of de phasing after RF pulse
• Frequency encoding gradient: applied at time of read out of signal

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
6
Q

What is the slice selection gradient?

A

• Applied during RF pulse
If add gradient stronger toward head, field strength = sum of gradient and B0 so nuclei higher up spin faster than those in weaker field so if apply specific larmor field, can stim the p+ in a specific slice to image it
• RF pulse only affects nuclei that experience a total field strength with matching Larmor frequency
• Slice : volume of excited nuclei
• One slice per RF pulse if 2D image –> scanning a full 3D image requires as many RF pulses as number of slices needed
• Interleaved slice acquisition: to minimize cumulative effects due to cross slice excitation
• The excited nuclei (the slice) are then affected by the other 2 gradients

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
7
Q

What is the phase encoding gradient?

A

Field is stronger in one row than the next so phases are diff bc row with gradient spins faster between row and if turn off again spin at same speed again but cant go back to same phase
• Applied after RF pulse
• Change spin resonance frequency of excited nuclei depending on their location in the gradient, causing de phasing
• When removed, resonance frequencies are the same again, but differences in phase persist
• All nuclei at a certain position in the gradient have same phase, thus phase is informative about position

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
8
Q

What is the frequency encoding gradient?

A

• Applied during data acquisition
• = the “read out gradient”
Creates slope in mag field in x direction, makes p+ in diff columns spin at diff direction
• All nuclei at a certain position in gradient have same resonance frequency, thus frequency at read out is informative about position

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
9
Q

What are pulse sequences?

A

• Pulse sequence: succession of RF pulses and gradient changes
• Example: Gradient echo echo planar imaging (GE EPI)
• Echoes are elicited by gradient reversals

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
10
Q

How do we create a spatial image from the signal?

A

• From one slice (slice selection gradient)
• Phase encoding gradient: atoms in different rows = different accumulated phase shift
• Frequency encoding gradient: atoms in different columns = different frequency during read out
• Sufficient echo’s –> all combinations of phases and frequencies are characterized
• To reconstruct an image the MRI signal is analyzed with frequency decomposition techniques Fourier analysis)

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
11
Q

What does a Fourier analysis do?

A

• Acquired temporal MRI signal decomposed into frequency components
• Spectrum expressed in polar coordinates of frequency and orientation (k space)
• Inverse Fourier transformation to create a spatial image based on these spectra

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
12
Q

What happens if theres an artefact in the k-space?

A

If theres an artefact in the k space (very bright spot somewhere), it would affect the whole image and would get stripes in the whole image, if bright spot was further away (high f noise) we would get dense stripes

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
13
Q

What is a voxel?

A

• Unit of space = voxel
• The shorter the time in which an image has to be taken, the lower the number of slices that can be imaged
• Number of voxels per row/column in the slice relates back to number of steps of phase encoding gradient

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

How do we get noise and imperfections?

A

• Related to imperfections in the magnetic fields, used pulse sequence and to Fourier spectrum analysis
• E.g. stripes or spikes (often reflecting point in k space where something went wrong)
• E.g. ghosting (presence of reflections/shadows of actual anatomy)
• E.g. geometric distortions like stretching and shearing

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
14
Q

How do we get noise and imperfections?

A

• Related to imperfections in the magnetic fields, used pulse sequence and to Fourier spectrum analysis
• E.g. stripes or spikes (often reflecting point in k space where something went wrong)
• E.g. ghosting (presence of reflections/shadows of actual anatomy)
• E.g. geometric distortions like stretching and shearing

How well did you know this?
1
Not at all
2
3
4
5
Perfectly
15
Q

How do we get contrast between tissue types?

A

We can see diff structures bc of diff brightnesses
Brightness of a voxel depends on strength of signal going there
Strength of signal correlates with # of p+ in that space
• Emitted signal decays over time
• Signal intensity depends upon several factors:
• Proton density
• T1 recovery : recovery of longitudinal orientation = spin lattice relaxation
• T2 decay : loss of transverse magnetization due to the loss in phase coherence = spin spin interactions
• Factors are different in different tissues, resulting in signal contrast

16
Q

What is T1 recovery?

A

• Recovery spans a relatively long time
• T1 contrast influenced by the time between successive excitations, the repetition time (TR)
• E.g. with 1 second, no tissue except fat will have mostly recovered but CSF will only have recovered a bit
the choice of TR will depend on how much T1 contrast we get
Long TR will allow everything to recover so wont see much contrast, short TR will max diff in T1

17
Q

What is T2 decay?

A

• T2 decay occurs much faster than T1 faster than T1 recovery
• Depends upon time interval between excitation (or refocusing by a gradient switch) and data acquisition = echo time (TE)
Signals decay faster in fat bc p+ change phase faster but in CSF p+ stay in same phase for longer

18
Q

What are weighted contrasts?

A

• Pulse sequence and parameters determine the contrast (proton density, T1 recovery or T2 decay)
• Regime in which one factor gives strongest difference in magnetization between different tissues + in which effect of other factors is minimized
• Maximal T1 and minimal T2 differences between CSF and fat: an intermediate TR (big difference in T1 recovery) and a very small TE (hardly any T2 decay)

19
Q

What are 3 important weighted contrasts?

A

• T1 weighted: intermediate TR, short TE
• Fat and white matter (WM) = bright (fast recovery), grey matter (GM) = dark, CSF & other fluids = black

• T2 weighted: long TR, intermediate TE
• Fat and WM= dark (rapid de phasing), GM = less dark, CSF & other fluids = white
• Sometimes with reduction of signal from fluids: FLAIR: FLuid Attenuated Inversion Recovery

• Proton density imaging: very long TR, very short TE

20
Q

What are 10 parameters chosen by the user?

A
  1. Coverage: whole brain or specific region?
  2. Orientation of slices: coronal, horizontal, oblique
  3. Number of slices (a 3D volume = a set of 2D planes or slices)
  4. Slice thickness and inter slice gap (gap gives more coverage and less RF interference)
  5. Field of view (FOV): spatial extent of each dimension in a slice
  6. Matrix size: number of voxels in each dimension of a slice
  7. Spatial resolution: in plane (or in slice) voxel size = FOV / matrix size - The smaller the voxel, the lower the signal to noise
  8. Repetition Time (TR): time between RF pulses
  9. Echo Time (TE): time between RF pulse and read out
  10. Flip Angle (<= 90)