Magnetic Resonance Imaging

Question 1

How is MRI often “characterised”? (2; positive characteristics)

Answer 1

By its high soft tissue contrast and high spatial resolution

Question 2

How can you better see certain structures in MRI once the scans are obtained?

Answer 2

Tuning of contrast by adjusting MRI sequences; E.g to see grey matter, white matter, cerebrospinal fluid or blood better

Question 3

What allows us to tune these different contrasts?

Answer 3

We use different MRI sequences to produce these images

Question 4

How is MRI clinically relevant?

Answer 4

It’s v sensitive to pathological changes in the tissue structure; Shows either as hyper or hypo intense signals e.g strokes, tumour or plaques

Question 5

Comment on MRI based on its
1. Temporal resolution
2. Spatial resolution
3. Depth of Perception
4. Sensitivity

Answer 5

1. Temporal resolution: Minutes-Hours (Slow)
2. Spatial resolution: 25- 100um (pre-clinical); 1mm (Clinical) (V Accurate)
3. Depth of Perception: Limitless (Can look everywhere in the body with the same sensitivity)
4. Sensitivity: 10^-3 to 10^-5 M

Question 6

What is an advantage of MRI over other modalities like PET? (Apart from high tissue contrast)

Answer 6

-Non-invasive (except if contrast agent is used)
-Non-radioactive
-Signal originates from body itself

Question 7

Apart from anatomical images, how else is MRI useful?

Answer 7

We can measure how fibers run through the brain (DTI), what brain regions are active during tasks (fMRI)

Question 8

What is inside the MRI apparatus? Aka what causes the magnetic field?

Answer 8

A big coil (Looping wires). A big current is put through these wires when the MRI is installed. This causes a large magnetic field running through the coil. Once the magnetic field is high enough, they close the loop so that the current runs through the loop indefinitely. This is only possible because the coil is cooled with liquid nitrogen to almost 0 Kelvin; the absolute 0 temperature point.

Question 9

How does the MRI magnetic field compare to earth’s magnetic field?

Answer 9

Earths magnetic field is 25 uT (0.000025 T), the MRI fields can be 1.5T, 3T or even 7T. Almost a factor 100,000 bigger

Question 10

How do you switch off the magnetic field?

Answer 10

The magnetic field is always on! Take off all metal

Question 11

What in our body can produce a signal to make images using a magnet? What characteristics allow for this?

Answer 11

Comes from hydrogen atoms (handy since body consists mainly of water and lipids.) Hydrogen atoms can produce an MRI signal because they have magnetic properties and angular momentum. Their spin and their charge create a magnetic moment and their spin and mass create angular momentum.

Question 12

Explain the concept of angular momentum

Answer 12

A gyroscope will keep spinning and not fall down because of gravity, instead it rotate around the direction of the gravitational force due to its angular momentum. In the same way, the magnetic moment of the hydrogen atom rotates around the direction of the magnetic field in the MRI scanner.

Question 13

Can we know the exact speed of this spin?

Answer 13

We know the exact speed of the rotation; it’s the product of the magnetic field strength of the MRI scanner and the ‘gyromagnetic ratio.’ Which is specific to the type of atom: Hydrogen is 42.57 MHz/T. This means that for 1.5T it rotates 60 Hz (spins 60 million rounds per second), for 3T 120MHz and 7T 300MHz.

Question 14

Apart from this fast spin precession, why else do we get lots of magnetic moments?

Answer 14

One small volume of tissue can contain many hydrogen atoms. This means may magnetic moments. One cubic mm of tissue contains around 3.35 x 10^19 water molecules.

Question 15

If were to graph the magnetic moments in a piece of tissue over time how would it look? Does this change in the MRI scanner?

Answer 15

All moments pointing in random directions and actually continuously moving over time. Magnetic moments preferentially align with the main magnetic field (B0) however and when we place our body in the MRI scanner. Thermal energy is far too high to make them actually align with B0 yet, a stable/ static et magnetic moment is formed (M0.) Think of compasses in a washing machine, the movement makes it too difficult to point north but if you took the average of the vectors, it would be pointing north.

Question 16

How do some textbooks depict the magnetic moments which causes confusion?

Answer 16

They state that the moments can only be parallel or anti-parallel with the magnetic field. This is not only wrong but will cause a lot of confusion understanding even the most basic MRI experiments.

Question 17

Are we more concerned with the net magnetic moments or the separate magnetic moments? Why?

Answer 17

We only care about this net magnetisation M, this makes our understanding of MRI much easier. Nearly all clinically relevant MRI studies can be understood by looking at the net magnetisation factor M.

Question 18

But how can M be turned into an MRI signal?

Answer 18

Radio frequency (RF) coils inside the MRI scanner carry out this function. The body RF coil produces a magnetic field (B1) perpendicular to B0. This causes an RF excitation of the tissue magnetisation which makes the magnetisation M precess (or rotate) around B0 and the RF field. In order to make this possible, the RF field needs to be on resonance with precessing M. The net magnetisation vector M also precesses with the longer frequency for the individual magnetic moments.

After we turn off the RF pulse, the tissue magnetisation will continue rotating around the main magnetic field B0. If we then place another coil around the tissue of interest, this rotating magnetisation can actually cause another electrical current through that coil. This is called induction.

Question 19

What does it mean to say that the RF magnetisation field needs to be on resonance with precessing M?

Answer 19

If the RF rotates too slowly then we can see that the magnetisation cannot be rotated away from its original position. If the RF field rotates too quickly, also nothing happens. Only if the RF field matches the longer frequency of the hydrogen atoms (120 MHz at 3T) the magnetisation will be rotated. We call this the on resonance condition, and the reason we say magnetic resonance condition. This resonance can be seen in many daily life things e.g pushing someone on a swing.

Question 20

Where do the waves that MRI utilise fall on what spectrum? To what extent are these damaging?

Answer 20

MRI uses radio frequent waves (RF) which fall between radio waves and microwaves on the electro magnetic spectrum. These are electromagnetic waves just as we use for regular radio signals, micro waves and even visible light. Since the frequency of the RF waves used in MRI are relatively low, they are harmless to the human body.

Question 21

How fast are these RF magnetisation pulses?

Answer 21

Very quick; they can rotate the magnetisation in around 0.2-2 ms.

Question 22

Based on the information so far, give a defiition for the MRI signal

Answer 22

The MRI signal is an electrical current through a receiver coil, caused by precession of tissue magnetisation around the main magnetic field B0.

Question 23

Why do we almost never depict the main precession (around B0)

Answer 23

It is too fast and although we know it is the source of our MRI signal, it is actually irrelevant in understanding most MRI processes.

Question 24

What do we do instead of depicting the main precession (around B0)?

Answer 24

We act as if the frame of reference is rotating with the lamar frequency (w0), then the magnetisation M and the RF field appear to stand still.

Question 25

Describe another characteristic/ parameter of the RF excitation

Answer 25

The flip angle (α) is how far the magnetisation is rotated. This most importantly depends on the product of the strength of the RF fields and the length of the RF fields:

(α = γ B1 △t)

The longer you keep the RF field on, the larger the rotation of the magnetisation will be.

Question 26

What is the ideal flip angle and why?

Answer 26

Ideally you want to rotate the magnetisation at an angle of 90 degrees. This is because at any point in time you can think of M being composed of 2 vectors: the longitudinal magnetisation pointing in the direction of the magnetic field and is therefore static, the transverse magnetisation on the other hand always lies on the transverse plane. Only the transverse magnetisation precesses around B0 and is therefore able to produce the MRI signal. Using an RF pulse of 90 degrees gives the highest amplitude of the transverse magnetisation and therefore produces the highest MRI signal.

Question 27

Where do the MRI coils need to be located in order to produce the MRI signal?

Answer 27

The MRI coils need to be located near the tissue of interest in order to sense the flipping of the tissue magnetisation. For this reason we have different coils for different body parts. It is also possible to perform RF excitation and reception with these kind of coils (1 coil).

Question 28

This means that all the tissue magnetisation seen by the coil are converted into one electrical signal. If signal comes from coil how can we know which tissue is where?

Answer 28

This function is carried out by gradient coils inside the scanner. Gradients in the scanner produce spatially varying magnetic fields.

Question 29

In what direction can the gradients be applied?

Answer 29

Gradients can be applied independently in x, y and z-direction.

Question 30

What causes the sounds associated with MRI?

Answer 30

Switching of the gradients

Question 31

What does this switching of gradients mean for a 3T MRI scanner?

Answer 31

For instance, they can make the field slightly lower at one end of the scanner (2.9T) and slightly higher at the other end of the scanner (3.1T).

Question 32

Why is switching the gradients useful?

Answer 32

Remember the field strength determines the frequency with which the magnetic moment rotates around the main magnetic field which will be different along each direction of the gradient. The higher end will have a slightly higher frequency signal than the lower end and their relatively coherent signals (in terms of amplitude and frequency) will come together to form a more complex signal which is the actual measured MRI signal. Since we know which gradient we applied, we actually know which frequency would be present at each location in the scanner. We can then recover the different frequency signals and their amplitudes from the total MRI signals.

Question 33

How can we recover the different frequency signals and their amplitudes from the total MRI signals?

Answer 33

We can do this mathematically through a trick called a fourier reconstruction. A fourier reconstruction decomposed any signal into its separate frequencies and amplitudes. Think in terms of our ears; music produces a one dimensional longitudinal wave in our ears similar to an MRI signal, yet our ears are capable of distinguishing all kinds of different sounds with different high and low frequencies from this sound wave.

Similarly, for each frequency we know its location in the scanner because of the gradient. If we know the amplitude then we can assign that amplitude to its corresponding location.

Question 33

How can we recover the different frequency signals and their amplitudes from the total MRI signals?

Answer 33

We can do this mathematically through a trick called a fourier reconstruction. A fourier reconstruction decomposed any signal into its separate frequencies and amplitudes. Think in terms of our ears; music produces a one dimensional longitudinal wave in our ears similar to an MRI signal, yet our ears are capable of distinguishing all kinds of different sounds with different high and low frequencies from this sound wave.

Similarly, for each frequency we know its location in the scanner because of the gradient. If we know the amplitude then we can assign that amplitude to its corresponding location.

Question 34

What is the relationship between transmit coils and receive coils?

Answer 34

Radiofrequent magnetic fields (from transmit coils) excite magnetisation and make them precess. Precessing magnetisation produces the MRI signal in receive coils.

Question 35

What does NOT determine the size of the net magnetization vector?
• The number of protons (1H) in the tissue
• The strength of the main magnetic field (B0)
• The temperature
• The gyromagnetic ratio

Answer 35

The gyromagentic ratio (a constant for any given nucleus that relates the nuclear MR frequency and the strength of the external magnetic field); the higher the temperature the more energy the atoms have to go against the field.

Question 36

What does NOT determine the signal strength after the RF pulse?
• The size of the net magnetisation vector
• The flip angle of the RF pulse
• The gyromagnetic ratio
• None of the above

Answer 36

None of the above; the higher the ratio, the faster the atoms spin, the higher the current generated by the atom.

Question 37

What do we need to change to perform MRI of atoms different from 1H?
• Use a different field strength
• Change the frequency of the RF pulse
• Change the length of the RF pulse
• All of the above

Answer 37

Change the frequency of the RF pulse; The rotation of magnetisation happens when the RF pulls it at the right frequency (due to different gyromagnetic frequencies)

Question 38

Describe some of the basic timing parameters in MRI

Answer 38

Repetition time (TR [sec]): sometimes multiple RF pulses are needed to make an image.
Echo Time (TE [me]): Time we decide to look at the magnetisation

Question 39

What can repetition time be for?

Answer 39

Repitition time can be for:
-spatial encoding steps (different gradient settings)
-averaging (If the signal is not high enough yet)
-dynamics
-…