MRI

Question 1

What are 3 advantages of MRI?

Answer 1

Is non-invasive
Is non-destructive
Uses no ionising radiation, making serial studies more ethically acceptable

Question 2

what are 4 disadvantages of MRI?

Answer 2

Time - takes longer. therefore patient motion a greater problem
Contraindications - More patients have contraindications for MRI compared to CT. Such contraindications arise because of metal in the patient in the form of passive medical implants or metallic fragments in their eyes, or because of active electronic implants such as cardiac pacemakers
Cost - expensive
Difficult theory

Question 3

will muscle produce a higher signal than cortical bone?

Answer 3

Yes

Question 4

how is it that the presence of bone can be inferred from MR images?

Answer 4

(1) cortical bone shows up as dark against the high signal from soft tissue and (2) marrow (i.e. fat) in the trabecular bone also produces high signal (Fig 1).

Question 5

Very briefly how does MRI work?

Answer 5

(1) place the patient in a strong external magnetic field to align the protons; (2) apply a pulse of RF radiation at the correct frequency to produce resonance; (3) detect the RF radiation emitted as resonating protons relax to give an NMR signal

Question 6

What quantity is the static magnetic field and how do we refer to it?

Answer 6

The static magnetic field (B0) is a vector quantity (i.e. it has direction as well as magnitude)

Question 7

What effect does B0 have on the patient?

Answer 7

exerts a force on hydrogen nuclei (protons) within the body and this causes them to align themselves in the direction of the field or in the opposite direction.

Question 8

What does the RF pulse do?

Answer 8

The RF pulse causes the nuclei to change their alignment but only if the radio waves within it have a particular frequency (the Larmor frequency).

Question 9

What happens after the RF pulse is switched off?

Answer 9

After the pulse is switched off, the patient emits RF radiation at the same frequency as the protons revert (relax) to their original alignments in the magnetic field. This emission is detected by a receiver to create a measurable signal (an electrical voltage that varies with time).

Question 10

How is the strong static magnetic field created?

Answer 10

by the flow of direct current in coils of electrically conducting material

Question 11

Why is the static magnetic field always on and present?

Answer 11

In the large majority of MRI scanners, the coil material is maintained at a very low temperature (that of liquid helium) such that it is superconducting i.e. it has no electrical resistance. For practical reasons, this means that the current is always on and that the field is always present.

Question 12

How is the RF pulse created?

Answer 12

The RF pulse (often called the RF magnetic field) is created by the flow of alternating current in a separate coil (that is not superconducting). The same coil may be used to detect the emitted RF radiation at a later time, or a separate coil may be used for this purpose.

Question 13

What are gradient coils and why do we have them?

Answer 13

The MRI scanner also has a further set of three non-superconducting coils (the gradient coils) that are used to locate the source of the emitted RF radiation within the patient.

In these gradient coils, direct current is rapidly switched on and off to produce gradient magnetic fields in three mutually perpendicular directions (X, Y and Z).

Question 14

What does ‘proton’ refer to in MRI?

Answer 14

the nucleus of a hydrogen atom (hydrogen-1) and does not refer to a proton in the nucleus of atoms of other elements that exist in the body

Question 15

What are 4 key pieces of information about the proton to understand MRI?

Answer 15

Protons have mass (1.7 × 10-27 kg), positive electric charge (1.6 × 10-19 coulomb) and ‘spin’
Spinning charges produce a magnetic field
Hydrogen nuclei therefore act like tiny magnets or magnetic dipoles (which have ‘north’ and ‘south’ magnetic poles separated by a short distance)
The human body is about 80% water and therefore contains lots of hydrogen nuclei

Question 16

What property does spin give to a proton?

Answer 16

angular momentum

Question 17

What is angular momentum?

Answer 17

loosely defined as the quantity of rotation possessed by an object. Angular momentum is a vector quantity

Question 18

How is angular momentum quantised on an atomic scale and what is it in a proton?

Answer 18

it can only have certain discrete values. The angular momentum of a nucleus is Ih/2π,where I is a quantum number that can only be zero, an integer (whole number such as 1, 2 etc.) or a multiple of ½ (such as ½, ³⁄₂, ⁵⁄₂ etc.). For the proton, I=½

Question 19

What is magnetic dipole moment?

Answer 19

The MDM is an important property of a magnet; it can be considered as the characteristic of the magnet that indicates how quickly it will align itself with an external magnetic field. The MDM is also a vector quantity.

Question 20

How do you calculate MDM?

Answer 20

the product of the spin angular momentum and γ, the gyromagnetic ratio

Question 21

What does the gyromagnetic ratio depend on?

Answer 21

γ is the MDM divided by the spin angular momentum. Its value depends on the type of nucleus; for the proton it is 2.67 x 108 rad s^-1T^-1. Here ‘rad’ means radian, the SI unit of angle, with 2π radians (a full circle) being equal to 360°.

Question 22

What happens to the protons in a body when in a high magnetic field?

Answer 22

protons align either with the field (in a relatively low-energy state) or against the field, i.e. in the opposite direction to it (in a higher-energy state). A slightly greater number of hydrogen nuclei align with the field than against the field

Question 23

What can happen to a proton in a magnetic field if applied with an oscillating magnetic field?

Answer 23

An oscillating magnetic field at the correct frequency can make protons change from the low- to the high-energy state and in the opposite direction. An oscillating field of this type is equivalent to electromagnetic radiation at the same frequency

Question 24

What is the equation for Larmor frequency?

Answer 24

ν = γB0/2π

Question 25

Why does a spinning proton create a magnetic field?

Answer 25

A moving electric charge produces a magnetic field, and so the proton may be regarded as a tiny magnet because it is a charged particle that is spinning rapidly about its axis

Question 26

Does the proton align exactly with B0?

Answer 26

No - a quantum mechanical rule prevents exact alignment and its MDM μ is inclined at an angle to the external field B0, which causes it to behave in a rather different way.

Question 27

What is a precessional orbit?

Answer 27

the proton also has spin angular momentum and this combination of spin and torque causes μ to precess around the direction of B0. Precession is a type of rotational motion that is completely different from the spinning of the proton about its own axis. During precession, the proton MDMs travel on the surface of two cones whose axes are in the direction of the external magnetic field.

Question 28

What is precessional frequency?

Answer 28

The rate at which the gyroscope precesses about the direction of the gravitational field G

Question 29

What is v in the larmor equation?

Answer 29

proton/Larmor precessional frequency -When the gyromagnetic ratio and the magnetic field strength are expressed in their SI units, the unit of the Larmor frequency is the hertz; it expresses the number of precessional rotations per second.

Question 30

What is the larmor processional frequency directly related to?

Answer 30

The Larmor precessional frequency is directly proportional to the strength of the magnetic field.

Question 31

What is magnetisation?

Answer 31

Magnetisation is the net MDM density (MDM per unit volume) and it is a property of the material (in this case human tissue), whereas MDM is a property of the magnet (in this case proton).

Question 32

How can the MDM vector μ be decomposed further?

Answer 32

μ2 in the direction of B0 and a component μ1 perpendicular to B0

the μ2 components generate a longitudinal magnetisation M along the direction of B0

the μ1 components are in random phases (they do not have phase coherence) and they are rotating due to precession (dashed arrows) - they cancel out and so there is zero transverse magnetisation in the plane perpendicular to B0

Question 33

How do you increase NMR signal strength?

Answer 33

increase magnetisation
the only practical way to do this is increase the size of the static field - ie increase the T of the coil

Question 34

What is the net magnetisation?

Answer 34

Their net magnetisation (M) is in the direction of the main field and is sometimes called the ‘equilibrium magnetisation’. The amount of magnetisation in the direction of the main magnetic field (the B0 direction) is 100% in this equilibrium condition and so we say the spins have 100% longitudinal magnetisation

Question 35

Why can’t we measure M in the z axis/

Answer 35

we cannot measure longitudinal magnetisation directly because M is in the same direction as B0 (the Z direction) and many orders of magnitude smaller than B0

Question 36

What is excitation?

Answer 36

tip over the magnetisation by typically 90° to convert it into transverse magnetisation and then measure that. This is called excitation.

Question 37

What is resonance?

Answer 37

We can tip or ‘flip’ the magnetisation over by applying a rapidly oscillating magnetic field at 90° to the B0 field (i.e. in the X or Y directions). However, the varying field will only flip the spins if it oscillates at the same frequency as the precession of the spins (i.e. the Larmor precessional frequency); this is what we mean by ‘resonance’.

Question 38

What is the RF field normally referred to as?

Answer 38

B1

Question 39

What trajectory does the magnetisation follow when in B1 field?

Answer 39

can be thought of as alternately pushing and pulling at the magnetisation, progressively tilting it away from the B0 direction. The magnetisation follows a spiral trajectory until it is precessing (or rotating) in the plane at right angles to the B0 field (in the transverse plane (the XY plane))

Question 40

What is the loss of transverse magnestisation called?

Answer 40

T2 relaxation

Question 41

T or F:
Immediately after a 90° pulse-
A. 100% of equilibrium magnetisation is recovered
B. Longitudinal magnetisation falls to 0%
C. Transverse magnetisation falls to 0%
D. Transverse magnetisation increases to 100%

Answer 41

A. F
B. T
C. F
D. T

Question 42

What are the 5 steps of relaxation after a 90 degree RF pulse?

Answer 42

1. The longitudinal equilibrium magnetisation has now been flipped into the transverse plane. Immediately after the RF pulse, all the spins that contribute to the magnetisation are precessing at the same frequency and are in phase
2. Some spins immediately start to precess more slowly than others - The spins are no longer in phase. This causes the transverse magnetisation to decrease.
3. As time goes on, there is a greater and greater phase dispersion of the transverse components of the spin MDMs and the transverse magnetisation continues to decrease
4. Finally, there is no net phase coherence and therefore no net magnetisation in the transverse plane
5. At the same time as the spins precess and lose phase coherence, they also begin to reorient themselves along the longitudinal (Z) direction. Eventually the full longitudinal magnetisation is recovered, i.e. the spins return to their equilibrium condition.

Question 43

What happens to the received magnetisation signal after 90 degree RF pulse?

Answer 43

Immediately after the 90° pulse there is 100% transverse magnetisation. This high-amplitude magnetic field that is precessing (rotating) at the Larmor frequency will induce a high-amplitude oscillating voltage in the receiver coil at the same frequency.

However, as phase coherence is lost, the strength of this transverse magnetisation will diminish and so too will the high amplitude of the induced voltage in the receiver coil.

Question 44

What is free induction decay?

Answer 44

the high-amplitude oscillating voltage in the receiver coil at the same frequency. Varying voltage signals such as the FID are the raw data from which all MR images are reconstructed.

Question 45

Why do spins dephase?

Answer 45

each hydrogen nucleus itself acts as a tiny magnet because of its electric charge and spin. It therefore produces a tiny magnetic field which will affect the field experienced by adjacent nuclei. This tiny field may add to, or subtract from, the field produced by the MRI scanner and this changes the larmor frequency of the proton.

Question 46

What is spin-spin relaxation time?

Answer 46

spin-spin relaxation time is the time constant of the FID fall-off. Also called the transverse relaxation time or T2

Question 47

What is T2* and why does this occur?

Answer 47

The real situation, in which the envelope of the FID is a steeper decreasing exponential curve with time constant T2*. The reason for this is imperfections in the static B0 field produced by the MRI scanner.

Question 48

How do we overcome the effects of T2*?

Answer 48

by using spin echo sequences

Question 49

How does a spin echo sequence work?

Answer 49

Stage 1: immediately after the 90° pulse the spins start to dephase due to spin-spin interaction and the field imperfections
Stage 2: some spins therefore gain phase compared to others and so the phase angle between them increases.
Stage 3: A second RF pulse is applied. this is a 180° pulse.
Stage 4: this has the effect of flipping the whole transverse plane through 180°. The spins are still precessing counter-clockwise, but now the slower precessing spins have gained phase compared to the faster spins
Stage 5: at a certain time referred to as TE (the echo time, or time to echo) the two spins will be back in phase creating a detectable signal (the echo) whose amplitude is unaffected by main magnetic field inhomogeneities. and any loss of transverse magnetisation over the time period TE will be due to the ‘pure’ T2 of the spins.

Question 50

T or F:
Immediately after the 180° refocusing pulse -
A. Equilibrium magnetisation is fully restored
B. 100% transverse magnetisation is restored
C. The transverse magnetisation is greater than immediately after the 90° pulse
D. The transverse magnetisation is greater than at the echo time (TE)

Answer 50

A. F
B. F
C. F
D. F

Question 51

What is spin-lattice relaxation time?

Answer 51

recover of longitudinal magnetisation over time described by an increasing exponential curve that approaches a limiting value, the equilibrium magnetisation M, with a time constant T1, the spin-lattice relaxation time.

Question 52

Which is always bigger t1 or t2?

Answer 52

T1 is always greater than T2.

Question 53

How can we measure variance in T1 time?

Answer 53

we allow the longitudinal magnetisation to partially recover before repeating our spin echo sequence. The second spin echo converts the partially recovered longitudinal magnetisation into measurable transverse magnetisation. If, for example, the longitudinal magnetisation has recovered by only 45% of its equilibrium value, this will be converted to 45% of maximum transverse magnetisation.

The time between the first 90° pulse and the second in the repeated sequence is called the repetition time (TR).

Question 54

What is the difference in lesions in T1 and T2?

Answer 54

images based on T2, lesions tend to appear brighter than normal tissues, whilst in images based on T1 they will appear darker.

Question 55

Why do we weight towards T1 or T2 rather than equally weighting?

Answer 55

if T1 and T2 effects contribute equally to the signal that lesions will be isointense with normal tissue and therefore undetectable.

Question 56

How do we weight to T1?

Answer 56

Reduce TR to maximise T1 contrast (i.e. TR about 500 ms)
Reduce TE to the lowest achievable on the MRI scanner in order to minimise T2 contributions.
The nomenclature for a typical T1W spin echo (SE) sequence would be: SE 500/5 ms (TR/TE)

Question 57

How do we weight to T2?

Answer 57

Increase TR to minimise T1 contribution. In an ideal world we could make this 15 seconds or more, by which time even CSF with its very long T1 of 3 seconds would have recovered nearly all its longitudinal magnetisation. However, as we will see in later sessions, we need to repeat these sequences perhaps many hundreds of times in order to build up image information. Generally speaking, patients do not stay still for more than about 8 minutes. With a ‘conventional’ SE pulse sequence, in order to obtain an image resolution of 256 pixels we need to make 256 repetitions; this means a maximum repetition time of about 2 seconds
Increase TE to maximise T2 contribution
A typical T2W spin echo sequence would have the nomenclature:

SE 2000/80 ms (TR/TE)

Question 58

How do we weight for proton density?

Answer 58

Increase TR to minimise T1 contributions (up to a practical limit of about 2000 ms in a conventional SE sequence)
Reduce TE as much as practicable to minimise T2 contribution (i.e. a TE of about 5 ms)
A PD SE sequence would therefore be:

SE 2000/5 ms (TR/TE)

Question 59

What does proton density sequences measure

Answer 59

an index of how many hydrogen nuclei there are per unit volume of tissue.

Question 60

Why are relaxation times for tumours normally higher than those of normal tissue?

Answer 60

because they usually contain more unbound water.

Question 61

Why can fat be a problem in MRI?

Answer 61

fat has an unusually short T1 but a reasonably high T2. This means that it produces high signal in both T1 and T2-weighted images. It is for this reason that a number of strategies have been developed to reduce or eliminate signal from fat as the high signal can otherwise obscure pathology.

Question 62

What makes a T1 long or short?

Answer 62

In a pure liquid, molecules are tumbling rapidly and freely:
Molecules have little time to interact with each other
T1 is long

In viscous liquids, water binds to less mobile macromolecules (this is the nearest to soft tissues in the body):
It is easier to transfer energy
T1 is short

In solids, molecules are relatively fixed, so there is a reduced chance of their coming close enough for an energy exchange:
T1 is long again

Question 63

What makes a T2 long or short?

Answer 63

In liquids, spins are tumbling end over end, so the magnetic fields they produce tend to even out, and therefore cause less perturbation to other spins:
There is little net change in local magnetic field
There is reduced spin-spin interaction
T2 is long

In solids, molecules are fixed. Therefore, the local field is fixed:
Local field variations are therefore significant
Spins dephase
T2 is short

Question 64

If an SE 15 000/1 ms sequence was used to image the brain, which of the following weightings would it most likely represent?

Answer 64

PD

Question 65

What are the 5 steps in acquiring a gradient-echo sequence?

Answer 65

1. RF pulse
2. Slice selection
3. phase encoding
4. frequency encoding
5. sequence repetition

Question 66

What is involved in slice selection?

Answer 66

To excite protons in a particular slice, it is necessary to apply the RF(read a full definition of this term) pulse in the presence of a slice-select gradient. Note that the area of the negative lobe of the gradient is equal to half the area of the positive lobe - this ensures that phase coherence is maintained within the image slice.

Question 67

What is involved in Phase encoding?

Answer 67

it is necessary to repeat the pulse sequence a number of times (typically 256 or 512) in order to build up an image with that equivalent number of data lines, or pixels. Each time the sequence is repeated, a different phase-encode gradient strength is used. For example, the strength of this gradient may commence with a large positive value and gradually be decremented each time the sequence is repeated until a large negative value is reached. Each ‘rung on the ladder’ represents a phase-encoding gradient of a particular strength, and this is shown schematically in the pulse sequence.

Question 68

What is involved in frequency encoding?

Answer 68

The frequency-encoding gradient is applied after each step of the phase-encode gradient. and it is during the application of the frequency-encode gradient that the MRI signal is ‘read’. In fact the frequency-encode gradient is often referred to as ‘read’ gradient. Again, the area of the negative lobe of the gradient is equal to half of the area of the positive lobe - this ensures that phase coherence is optimised at the central point of the frequency encoding

Question 69

What is the Echo time of a sequence?

Answer 69

The echo time (TE) is the time taken from the RF(read a full definition of this term) pulse to the MRI signal, or echo.

Question 70

What is the TR time?

Answer 70

The repetition time (TR) is the length of time between one RF(read a full definition of this term) excitation pulse and the next one.

Question 71

How can the length of time for a scan be calculated?

Answer 71

If the TR is known, then the length of time taken to complete the sequence can be calculated. The scan time is equal to the TR value (typically 500-3000 ms) multiplied by the number of phase encode steps required (typically 256 or 512).

Also, if more than one signal average is required then the scan time will increase. Two signal averages will take twice as long to acquire, three signal averages will take three times as long to acquire, and so on.

Question 72

In a spin echo sequence where is the 180 degree pulse positioned?

Answer 72

The 180° pulse is positioned exactly half way between the initial RF pulse and the time at which the signal is acquired, in other words the 180° pulse is applied at a time equal to TE/2.

Question 73

What is the problem with SE sequences?

Answer 73

generally take longer to implement than GE sequences.

Question 74

What is an inversion recovery sequence?

Answer 74

An IR pulse sequence is really very similar to a GE or an SE sequence, except that a 180° ‘inversion pulse’ is applied at the start of the sequence. The effect of this 180° pulse is to flip (invert) the equilibrium magnetisation in the z direction. As soon as this has occurred, T1 recovery processes begin. In the sequence, a delay time TI (time from inversion) is built in to allow the scanner operator to control the resulting contrast that is based on the T1 recovery rates of tissues. Following the application of an inversion pulse and an appropriate TI delay, a GE(read a full definition of this term) or an SE(read a full definition of this term) sequence is run.

Question 75

What is a STIR sequence?

Answer 75

Short Tau inversion recovery - Looking more closely at the exponential recovery of the longitudinal magnetisation for two example tissues, e.g. fat and water, it can be seen that the (short T1) fat magnetisation passes across the TI axis – known as the null point - earlier than the (long T1) water magnetisation. At this specific TI time (approximately 130 ms at 1.5 T) there will be no longitudinal fat magnetisation available for the subsequent sequence and the resulting images will contain no fat signal. This is known as a STIR sequence.

Question 76

What is a FLAIR sequence?

Answer 76

Fluid attenuation inversion recovery - water magnetisation passes across the TI axis (null point) later than the fat magnetisation. At this specific TI time (approximately 2500 ms at 1.5 T) there will be no longitudinal water magnetisation available for the subsequent sequence and the resulting images will contain no water signal.

Question 77

T or F Regarding a GE pulse sequence:
A. RF pulses are separated by a time interval TE
B. An echo signal is acquired when a frequency-encoding gradient is applied
C. One phase-encoding gradient is used
D. The slice-selection gradient has two positive lobes
E. TR is the rise time of the RF pulse

Answer 77

A. F. RF(read a full definition of this term) pulses are separated by a time interval TR(read a full definition of this term).

B. T. Because it accompanies signal acquisition, the frequency-encoding gradient is also called the read or read-out gradient.

C. F. Typically, 256 or 512 phase-encoding gradients are used to produce one image.

D. F. The slice-selection gradient has a positive and a negative lobe; the area of the negative lobe is equal to half that of the positive lobe to preserve phase coherence.

E. F. TR(read a full definition of this term) is the interval between successive RF(read a full definition of this term) pulses.

Question 78

T or F Regarding an SE pulse sequence: A. A 180° RF pulse is used to invert the longitudinal magnetisation
B. The image acquisition time is shorter than that for a GE sequence
C. An echo is formed after a time interval TI
D. TR is twice as long as TE
E. The image acquisition time is the product of TR and the number of phase-encoding gradients

Answer 78

A. False. The longitudinal magnetisation has already been tipped into the transverse plane by a 90° RFpulse; the 180° RFpulse re-phases the magnetisation in the transverse plane.

B.False. The use of a re-phasing RF pulse means that, in general, the acquisition time is longer than that of a GE sequence.

C.False. An echo is formed at a time TE after the 90° RF pulse and a time TE/2 after the 180° RF pulse.

D. False . TR and TE are set independently by the MRI scanner operator; it is possible that TR could be twice as long as TE but, in general, it is longer than 2TE.

E. Correct. With no image averaging, the acquisition time is TR multiplied by the number of phase-encoding gradients that are used.

Question 79

What are the metal related sources of hazard with MRI?

Answer 79

The ferromagnetic missile effect on extraneous metal

Migration/rotation of metal implants or fragments in the body

Current induction and heating of extraneous metal

Current induction and heating of implanted devices

Question 80

What measures can help to guard against the ferromagnetic missile effect in an MR unit?

Answer 80

All patients and staff must be checked (‘sifted’) for ferromagnetic material before entering the Controlled Access Area

In some centres all patients are required to change out of street clothes into gowns or surgical ‘greens’ in order to ensure no clips or pins or coins are hidden in their clothing

Question 81

What measures can help to guard against the migration/rotation of metal implants in an MR unit?

Answer 81

the ‘Safety Checklist’
to ascertain if the patient, and any attending staff member, have potentially hazardous implants or fragments of metal in their bodies.

Question 82

How can burns come about in MRI?

Answer 82

the resonant antenna effect, whereby the MR’s radiofrequency (RF) pulses set up a standing voltage wave in the metal if the length is equivalent to the half wavelength of the RF radiation.

If this happens, the tips of wires eg ecg/pulse oximetry can undergo rapid heating and burn the patients

Question 83

What are the risks with cardiac pacemakers in MRI fields?

Answer 83

cardiac pacemakers or implanted defibrillators can malfunction when exposed to the fields of MR.

The pacemaker, its leads and the myocardium itself form a single conductive circuit. If this system happens to ‘tune’ to the frequency of the RF radiation, currents will be induced in the circuit. This may cause the heart to contract. With the MR often firing scores of RF pulses a second, there will not be sufficient time between contractions and little blood will be pumped

The resonant antenna effect may be responsible, with induced currents in the tips of the wires heating up where they are in contact with the myocardium and thus causing burns and thermal shock

Batteries and/or implant casings may be ferromagnetic and so may undergo migration and torque

The electromagnetic fields may cause electronic circuitry to malfunction

If the circuitry controls a drug reservoir in an implanted pump then there might be the danger of the whole reservoir being rapidly discharged into the patient

Question 84

Who publishes guidlenes for MRI safety?

Answer 84

Medicines and Healthcare products Regulatory Agency (MHRA)

Question 85

What do the MHRA MRI safety guidelines include?

Answer 85

Setting up a Controlled Access Area

Designation of staff who may be involved with MR and their different training requirements

The use of a checklist or questionnaire for those who enter the high magnetic field environment

Writing Local Rules detailing safe working practices in the high magnetic field environment including the control of equipment

The safety of different types of medical implant

Procedures in the event of emergency situations

Question 86

What are the 2 different lines or zones within the MRI controlled areas?

Answer 86

the 5 gauss line - the boundary outside which it is considered safe for people with implants such as cardiac pacemakers.

MR Projectile Zone -In the MR Environment, the volume within the 3 mT magnetic field contour, where the risk of the ferromagnetic missile effect is considerable

Question 87

What is important to clarify with implanted devices for MRI safety?

Answer 87

important to establish the make and model of the implant, as well as the manufacturer. The MR safety of any implant should be established by consulting the manufacturer about the particular make and model and its MR compatibility or otherwise.

Question 88

Are orthopaedic implants MRI safe?

Answer 88

generally made of a sufficiently high quality surgical steel that they have no significant ferromagnetic component. Nevertheless heat can be generated by the RF field of the MRI scanner. For this reason, patients need to be carefully monitored throughout their time in the magnet room, and need to be warned to press the emergency ‘call’ button if they feel the slightest discomfort.

Question 89

Why can cerebral aneurysm clips be unsafe in MRI?

Answer 89

although “non-ferromagnetic” they undergo a sterilisation process which ay encourage the build-up of ferromagnetic ‘domains’ in the metal.

Question 90

WHat must happen if the patient arrests in MRI?

Answer 90

the patient must always be removed from the magnet room, and preferably from the Controlled Area, before being made accessible to the crash team. Often they may bring a ferromagnetic trolley carrying many ferromagnetic objects and devices.

Question 91

How do you turn off the magnetic field in an emergency?

Answer 91

Superconducting quench - If the coil temperature rises above the superconductivity threshold, the windings suddenly develop a finite resistance. The several-dozen amperes of circulating current passing through this elevated coil resistance create heat. This heat causes a sudden, explosive boil-off of liquid helium.

Question 92

What can be dangerous about quenching an MRI machine?

Answer 92

Quenches present certain hazards to staff. Though not toxic, the gaseous helium will displace oxygen and so there is the danger of asphyxia. The gases are also very cold. As such the area should be evacuated

Question 93

What are the three modes of operation for an MRI machine?

Answer 93

Normal mode, which encompasses routine procedures in which the risk to the individual is minimal

Controlled mode, in which the exposure is greater and imaging performance improved.

Research/experimental mode, for which exposure is usually greater still and needs to be restricted to prevent harmful effects.

Question 94

Why is the effect of B0 (static magnetic field) normally very low on humans?

Answer 94

tissues of the body are primarily diamagnetic in nature and this means they are only very weakly ‘magnetisable’. Effects in strong fields is still possible though

Question 95

What can humans experience in fields of over 2T?

Answer 95

At fields of 2T or greater, humans can experience certain sensations including vertigo, dizziness, nausea and a metallic taste in the mouth; sensitivity to these effects varies between individuals. The effects are usually associated with head motion and may be caused by currents being induced in the semi-circular canals as they move through the magnetic field, thereby ‘cutting’ the magnetic lines of force and so inducing electric currents.

Question 96

What effects can the gradient fields cause?

Answer 96

Though the gradients produce maximum fields that are not as strong as the static field, perhaps a fiftieth of the value, their rapidly-changing nature means that they will induce electrical fields and hence currents in conductive materials, such as the tissues of the human body.

it is possible for the induced currents to cause PNS and muscle stimulation. This usually manifests itself as muscular twitching. Cardiac stimulation or epileptic fits might possibly result.

Question 97

What is the minimum threshold for inducing PNS symptoms?

Answer 97

20 Ts-1 is regarded as the minimum threshold

Question 98

What are the effects of the RF wave on tissues in the body?

Answer 98

the effect of the RF is power dissipation in the tissues of the body and their subsequent heating. Heat generated in tissue is usually compensated by thermoregulation, where dilation of the vasculature increases blood flow and so allows the heat to be carried away and dissipated via the skin and exhaled breath.

Question 99

Which tissues in the body deal less well with increase in temperature and are heat sensitive?

Answer 99

These include the eyes, which have little blood flow and the testes which should be at a lower temperature than the rest of the body.

Question 100

What measures heat energy transfer to the patient in MR?

Answer 100

In MRI, the specific absorption rate (SAR) is used as a measure of the amount of energy per unit time (power) deposited per unit mass of tissue in the patient or subject by the RF field. The usual units are watt per kilogram (Wkg-1).

Question 101

What are the limits for temperature rise for the whole body with MR?

Answer 101

Normal - 0.5 for whole body
controlled - 1
research - 2

Question 102

In normal mode what are the max temperatures for the head, trunk and limbs?

Answer 102

head 38
trunk 39
limbs 40

Question 103

What are the patient whole body SAR limits with MR?

Answer 103

Normal 2
controlled 4
research >4

Question 104

What does the SAR depend on?

Answer 104

a number of factors including the static magnetic field strength, the type of pulse sequence, its timing characteristics, the number of repetitions, the number of slices and the output power in each RF pulse.

Question 105

How is SAR roughly estimated?

Answer 105

fire a few test pulses of a certain power and measure the returned signal. It will take the patient’s weight (which the operator must input before the scan starts) and calculates the energy absorption of the body.

Question 106

Why do patients wear ear guards for MRI?

Answer 106

MR systems operating at 1.5 T produce noise levels in the range 80-110 dB(A). Clinically significant temporary shifts in hearing threshold may occur at 85 dB(A) and above

Question 107

What are the staff MR exposure limits?

Answer 107

should not be exposed to static magnetic fields in excess of 2 T to the head and trunk, which is half the value for patients in normal operating mode, and 8 T to the hands and limbs

the change in field strength should not exceed 2 T in a time period of 3 s

Question 108

What are the public MR exposure limits?

Answer 108

the acute exposure limit for the direct effects of static fields is 400 mT to any part of the body

the MR Environment contains the 0.5 mT contour and so the field strength outside this region would be much less than the exposure limit.

Question 109

What are the staff SAR exposure limits?

Answer 109

the whole body SAR exposure limit is 0.4 Wkg-1 whereas the local SAR exposure limit for head and trunk is 10 Wkg-1 and that for limbs is 20 Wkg-1. The whole-body SAR is averaged over 30 minutes while the local SARs are averaged over 6 minutes and 10 g of tissue.

only workers very close to the magnet during scanning might be subject to exposures approaching the limits and heating is very unlikely to occur in staff outside the imaging volume

Question 110

Above what level must staff wear ear protection?

Answer 110

85 dB(A)

Question 111

How should MR be changed in pregnancy?

Answer 111

optimised pulse sequence that minimises RF radiation and acoustic noise exposure.

Where possible, gradients should be switched to reduced acoustic noise mode (e.g. ‘whisper’ or ‘soft tone’). Longer pulse repetition time, reduced image resolution, increased slice width and larger field of view all reduce the rate of change of gradient fields and thus the acoustic noise generated.

a pregnant member of staff should not remain in the MR Environment whilst scanning is underway, to reduce the possibility of foetal exposure to EMFs and, in particular, to acoustic noise

Question 112

Is Gadolinium toxic?

Answer 112

Yes - The safety of gadolinium (Gd) contrast agents depends on the stability of the chelate as gadolinium itself is very toxic. At least 9 serious anaphylactoid reactions and 1 death are known to have occurred.

Question 113

How is a T1W image obtained?

Answer 113

by carefully considering the recovery of the proton longitudinal magnetisation for each tissue. By changing the repitition time with knowledge of tissue T1 relaxation times this can be manipulated to derive different signal intensities from the two tissues and therefore optimise the contrast between them. eg short TR for contast between CSF (T1 recovery 4000ms, WM 700ms)

Question 114

What do we take as T1 recovery time for a tissue?

Answer 114

T1 is the time when the signal has reached 63% of its final (maximal) value.

Question 115

How is a T2W image obtained?

Answer 115

by carefully considering the decay of the proton transverse magnetisation for each tissue. Can choose a TE at a point in time where there is maximal difference between the transverse decay curves for each tissue, and have therefore introduced T2 weighting. A longer echo time allows tissues with shorter T2 decay times to lose signal whereas longer T2 tissues will still retain tissue

Question 116

What do we take as T2 recovery time for a tissue?

Answer 116

the T2 value for each tissue is defined as the time taken for the transverse magnetisation to decay to roughly 37% of its original value.

Question 117

What is T2 weighted images good for generally?

Answer 117

highlighting tissues with an elevated free-fluid content - often this is typical of pathological lesions such as tumours.

Question 118

What is T1 weighted images good for generally?

Answer 118

anatomical detail

Question 119

Why do we need a Short TE as well as short TR to get a T1W image?

Answer 119

T1 recovery and T2 decay processes are occurring simultaneously (but independently), and so it is important to consider the effect of each relaxation parameter in parallel. we also need to use a short TE in order to minimise any T2 weighting in the image.

Question 120

What do you get if you take images with a long TR and short TE?

Answer 120

the resulting image should contain very little T1 weighting or T2 weighting. In this situation, the contrast information is primarily based on the proton density of the tissues, and is therefore referred to as PDW imaging.

Question 121

What tissue has both the longest T1 and T2 times?

Answer 121

CSF

Question 122

What has the shortest T2 time?

Answer 122

muscle

Question 123

What has the shortest T1 time?

Answer 123

Fat

Question 124

When might a different flip angle be used?

Answer 124

In GE imaging The resulting contrast is heavily dependent on this choice of flip angle

Question 125

What happens when you use a small flip angle?

Answer 125

a large longitudinal magnetisation component is retained and this recovers back to equilibrium very quickly.

As a result, it is not really possible to observe T1-based differences in the recovery curves for different tissues

Question 126

In GE imaging what weighting does a short TR, short TE and small flip angle give you?

Answer 126

proton-density weighting

Question 127

What happens when you use a large flip angle?

Answer 127

only a small longitudinal magnetisation component is retained.

There is therefore more scope for differences in T1 recovery between the different tissues to become apparent and the resulting image will be T1 weighted

Question 128

In GE imaging what weighting does a short TR, long TE and small flip angle give you?

Answer 128

T2* weighting

Question 129

In GE imaging what weighting does a short TR, short TE and large flip angle give you?

Answer 129

T1 weighting

Question 130

What is the Ernst angle and what is it affected by?

Answer 130

the optimal flip angle to use (Ernst angle θ) is governed by a relationship between the sequence TR and the tissue T1 relaxation time. although the Ernst angle may provide maximum signal intensity on a SPGR sequence for a single tissue, it does not necessarily maximise the contrast between two different tissues.

Question 131

What is spoiled gradient echo sequence?

Answer 131

where any residual transverse magnetisation is removed at the end of each TR. Other names for this type of sequence include fast low-angle shot (FLASH) and T1-fast field echo (T1-FFE).

Question 132

How does Inversion recovery work?

Answer 132

inversion recovery consists of a 180° inversion pulse followed by a delay time TI before implementation of a conventional MRI sequence (usually an SE. The delay time TI is critical in determining the relative contrast, e.g. between tissues that contain fat and those that contain water. The contrast generated by inversion recovery is dependent on the TI value being carefully chosen in order that the longitudinal component of the magnetisation is selectively removed whilst other tissues still have significant longitudinal magnetisation component present to generate subsequent transverse magnetisation.

Question 133

When is STIR particularly useful?

Answer 133

musculoskeletal applications where removal of fatty tissue signal may be required (e.g. detection of bone bruises).

Question 134

What is FLAIR and what is it useful for?

Answer 134

if a TI time of approximately 2500 ms is chosen, then the resulting image will contain signal from fatty tissue but not from fluid-based tissue. FLAIR sequences are useful for highlighting plaques in multiple sclerosis brain examinations.

Question 135

Where does noise predominantly come from in MRI?

Answer 135

Noise in an MRI image primarily comes from the RF coil and receiver system and the patients themselves

Question 136

How can noise from external RF signals be minimised?

Answer 136

adequate RF shielding (e.g. the screened room or Faraday cage).

Question 137

How can noise from the RF coil and receiver system be minimised?

Answer 137

use of appropriately designed RF instrumentation

Question 138

What happens to contrast as size of the object is reduced?

Answer 138

As the size of an object is reduced, it will retain its full contrast until its size matches the image pixel size.

When it becomes smaller than the pixel size, its contrast decreases and eventually the contrast falls below the threshold for perception. Objects may still be perceived even though they are smaller than the pixel size; this is more likely for inherently bright (high contrast) objects than an inherently dark (low contrast) objects

Question 139

What does the contrast of an object fall when it gets smaller than a pixel?

Answer 139

the signal from it will be ‘averaged out’ over the background and the resulting pixel will appear as relatively less bright

Question 140

What effect will doubling the image resolution have on the image?

Answer 140

halving the pixel size but SNR reduced by a factor of 4

Question 141

What happens if you halve the thickness slice?

Answer 141

SNR reduced by factor of 2

Question 142

What happens if you reduce receiver bandwidth by factor 2?

Answer 142

SNR Increased by factor of 1.4 (√2)

Question 143

What happens if you double the number of signal averages (NEX)?

Answer 143

SNR Increased by factor of 1.4 (√2)

Question 144

What happens if you double the number of phase encode steps?

Answer 144

SNR Increased by factor of 1.4 (√2)

Question 145

What is the equation for Contrast-to-noise ratio?

Answer 145

CNR= (Sa – Sb)/N

Question 146

Why is slice thickness important between different size lesions?

Answer 146

If the image slice thickness is thicker than the lesion and the voxel containing the lesion also has a component from the surrounding tissue a proportion of the lesion pixel signal is due to surrounding tissue; this is the partial volume effect. As a consequence, contrast between the lesion and the tissue is slightly reduced

Question 147

What is a localiser image?

Answer 147

this might consist of 21 images in three planes in a short time of about 24 seconds. These images have low spatial resolution and a large field of view. They are equivalent to ‘scout views’ in CT and are used for positioning the slices in the main clinical protocol.

Question 148

Is it the case that in an SE sequence, transverse magnetisation dephasing is refocused by a 180° RF(read a full definition of this term) pulse, and this allows T2*W images to be obtained?

Answer 148

False - A 180° pulse is able to refocus the transverse magnetisation dephasing, but this will result in T2W(read a full definition of this term) (not T2*W) images.

Question 149

What does tomographic mean?

Answer 149

the image you see is taken from a thin slice through the object being scanned

Question 150

What is the relationship between pixel size, Field of view and Matrix size?

Answer 150

pixel size = FOV/matrix size

Question 151

How is possible to have different resolutions in horizontal and vertical directions?

Answer 151

using rectangular FOVs or a rectangular matrix.

Question 152

What is the partial volume effect?

Answer 152

if the slice is thick enough, such that two or more tissues are within the imaging slice, the pixel value will be an average of the contributing tissues within the slice. it will reduce contrast and make small objects difficult to see.

Question 153

What is a sinc pulse?

Answer 153

RF pulse with a particular shape - The sinc shape is that of the pulse envelope

Question 154

What are mixed together to make an RF pulse?

Answer 154

from an envelope function (the pulse shape) and a sine wave oscillating at the Larmor frequency.

Question 155

What is the difference between characteristic time and pulse duration?

Answer 155

Characteristic time is the width of some feature in the function (e.g. for a sinc function, this characteristic time is the width of the central lobe) whereas the pulse duration is the time for the whole envelope (theoretically infinite)

Question 156

What is bandwidth?

Answer 156

Because the carrier wave is shaped by the envelope function, the resulting RF pulse actually contains a range of frequencies, rather than just one at the Larmor frequency. This range of frequencies also has a shape and a characteristic frequency range (bandwidth)

Question 157

What is the relationship between pulse duration and bandwidth?

Answer 157

There is an inverse relationship between pulse duration and bandwidth. If one goes up, the other goes down and vice-versa.

Question 158

What is the difference between frequency and phase of a signal?

Answer 158

The frequency is defined as the number of peaks counted in some time interval. Phase is the position of the peaks relative to some reference, usually another wave.

Question 159

WHat happens to the MR signal of an object if a gradient is applied to it?

Answer 159

it decreases - the MR signal oscillates at the Larmor frequency. When a gradient is applied, the Larmor frequency of the MR signals from the apple will vary with position. The total MR signal picked up by the detector will decrease over time as the individual signals go out of phase with each other. The amount of dephasing will depend upon the gradient strength and its duration (effectively the area under the gradient on a pulse sequence diagram).

Question 160

How is a slice selected in MRI?

Answer 160

Gradient fields change the larmour frequency which vary with position in relation to the field. As such only spins at the larmour frequency will be in a narrow plane at right angles to the gradient.

Specific RF pulses at the larmour frequency will only excite those in the narrow field.

Question 161

How are different scanning planes achieved in MRI?

Answer 161

the Z gradient acts as a slice selection gradient and the slices are in the transverse (axial) plane. Using the X gradient as the slice selection gradient would give sagittal slices while using the Y gradient in this way would give coronal slices. Furthermore, it is possible to use a combination of gradients to produce slices in any plane; this flexibility is one of the advantages of MRI.

Question 162

What is the rephasing gradient?

Answer 162

At the end of the RF pulse, the MR signal needs to be put back into phase. This is achieved using a ‘rephasing gradient’, applied in the opposite sense to the gradient used during slice selection.

Question 163

How can slice width be affected?

Answer 163

by the pulse duration (i.e. bandwidth), Larmor frequency and the strength of the gradient

Question 164

What happens to slice width if you decrease bandwidth?

Answer 164

smaller slice

Question 165

What happens if you increase the gradient strength of the field?

Answer 165

smaller slice

Question 166

What happens if you increase larmor frequency?

Answer 166

move the slice Up (if in axial)

Question 167

T or F - Regarding the factors that affect slice properties:
A. Increasing the pulse duration increases the bandwidth
B. Increasing the pulse duration reduces the slice thickness
C. Reducing the gradient strength reduces the slice thickness
D. Decreasing the Larmor frequency of the carrier wave moves the slice to the left
E. The thinnest slices can be obtained using high gradient strengths and long RF pulses
F. For a slice that is off-centre, increasing the gradient strength has no effect on the slice position

Answer 167

A. False
B. True
C. False
D. True
E. True
F. False

Pulse duration is inversely proportional to bandwidth. Thus, increasing the pulse duration decreases the slice thickness. A stronger gradient (steeper line) reduces the slice thickness and changing the Larmor frequency will move the slice. Changing the gradient strength for an off-centre slice will alter both the slice thickness and position. On a clinical MRI scanner, the user would select the required slice thickness and position, and the computer would calculate the required RF and gradient parameters automatically.

Question 168

What properties are used to localise signal from within a slice?

Answer 168

utilising properties of the received signal, such as frequency and phase.
A read-out gradient is applied by acquiring the MR signal in the presence of a gradient, such that the Larmor frequency of the MR signal will depend upon position along the gradient.

Question 169

What is gradient echo?

Answer 169

One of the consequences of applying the gradient is that the MR signals dephase. There is a small but finite amount of time before the MR signal can be collected, and in this time, a large reduction in the signal can occur. To get around this, the MR signals are deliberately dephased using a gradient with the opposite sense to the read-out gradient.

This means the read-out gradient is rephasing the MR signals, such that at some point during the acquisition of the data, they all come back into phase and maximum total signal is detected. This is called a gradient-echo

Question 170

How are individual frequencies extracted from the total MR signal?

Answer 170

Fourier transform - The FT takes the data that has been acquired over time, extracts the frequencies and outputs a plot of the height (or amplitude) of the individual frequencies against frequency

Question 171

How is MR signal digitised?

Answer 171

sampling by an analogue-to digital converter (ADC). - ‘looks’ at the height of the MR signal at regular intervals at a given sampling time/sampling frequency

Question 172

How are sampling and pixels related?

Answer 172

The number of samples will equal the pixels on the final image.

Question 173

What is aliasing in MRI?

Answer 173

where high frequencies may appear as low frequencies due to sampling

Question 174

Why is aliasing an issue in MRI?

Answer 174

Since the MR signal frequency relates to spatial position, aliasing causes the MR signal to be assigned to the wrong position in the final image.

Question 175

what is the rule with sampling to reduce aliasing?

Answer 175

you must sample at twice the maximum frequency contained within the MR signal; this is called the Nyquist criterion.

Question 176

What is the nyquist frequency?

Answer 176

The maximum frequency that can be accurately digitised is called the Nyquist frequency; it is equal to half the sampling frequency.

Question 177

How is the nyquist criterion not broken?

Answer 177

the range of frequencies is limited prior to sampling, using something called a ‘band-pass’ filter. The frequency range the band-pass filter will allow through is called the receiver bandwidth.

Question 178

How is the field of view set in MRI?

Answer 178

The FOV in the frequency-encoding direction will be defined by the receiver bandwidth range of frequencies

Question 179

How is spatial localisation performed in Y axis?

Answer 179

utilise the phase of the MR signal. Done by applying a phase-encoding gradient for a short time before the readout period. The short application of gradient changes the oscillation speed for a period and then they oscillate at the same speed again but at different phases after gradient turned off

Question 180

Why are multiple phase encoding gradients required in practise?

Answer 180

A single phase-encoding gradient is not sufficient to determine the vertical position of the MR signal. Measuring the phase is very difficult since it is not absolute, but relative to some reference. There are also other factors (e.g. inhomogeneities in the main magnetic field) that can also change the phase of the MR signal, resulting in positional errors

Question 181

What happens to the multiple phase encoding signals?

Answer 181

frequencies can be extracted using an FT in the same way as that done for frequency encoding.

Question 182

How can aliasing effect phase encoding?

Answer 182

waves from regions outside the FOV produces an identical wave to those inside FOV Thus, signal from outside the FOV aliases back into the FOV on the other side of the image.

Question 183

What is phase wrap around?

Answer 183

where some anatomy at the edge of the FOV appears on the other side of the image eg a nose at the back of the head.

Question 184

How can you reduce pahse encoding aliasing?

Answer 184

it may be eliminated by increasing the FOV in the phase-encoding direction, at the cost of poorer spatial resolution.

Question 185

In MRI raw data what information is encoded in each row and column?

Answer 185

Each row contains the data obtained during frequency encoding. Successive rows are filled by repeating the MR acquisition with different phase-encoding gradient strengths. The phase-encoding gradient is usually stepped from a large negative value through zero to a large positive value.

A column in the raw data represents a single frequency-encoding sampling point at different phase-encoding steps. Thus, a vertical column is the sum of the phase change waves built up by successive phase-encoding steps.

Question 186

What is k space?

Answer 186

k-space is a concept used to describe positions in the raw data array; indeed, it can be considered as the raw data array itself.

Question 187

How is k-space related to the final image?

Answer 187

via the FT

Question 188

What does the centre of the k space encode?

Answer 188

the contrast information for the final image (i.e. it defines the regions of light and dark).

Question 189

What does the outer part of the k space encode?

Answer 189

correspond to the raw data that contains the detail information i.e. defines the boundaries between tissues.

Question 190

How can the FOV in the frequency encoding direction be changed?

Answer 190

defined by the maximum and minimum frequencies contained within the MR signal. This is defined by the receiver bandwidth and translated into the FOV via the gradient.

Question 191

What does changing matrix size change?

Answer 191

simply involves taking more data samples. However, this will extend the time of data acquisition, which may change other parameters such as the repetition time (TR) and echo time (TE).

Question 192

Say you wanted to double the matrix size in the frequency-encode direction while keeping the MR signal acquisition time and FOV constant. What would you need to change?
A. Receive bandwidth only
B. Frequency-encoding gradient strength only
C. Both receive bandwidth and frequency-encoding gradient strength

Answer 192

The correct answer is C.

You should change both receive bandwidth and frequency-encoding gradient strength.

In order to acquire twice as many data samples in the same time, you need to halve the time between samples. Therefore, the bandwidth must be doubled (recall that bandwidth is related to the inverse of the time between samples). However, doubling the bandwidth will double the FOV; to keep the FOV constant, the frequency-encoding gradient strength needs to be doubled.

Question 193

How is FOV controlled in phase encoding direction?

Answer 193

the phase-encoding increment (i.e. the increase in the area under the gradient between successive phase-encoding steps). Increasing the phase-encoding increment will have the effect of reducing the FOV. However, this may result in phase aliasing.

Question 194

How can phase-encoding increment be changed?

Answer 194

either increasing the gradient strength or the gradient duration as the phase shift depends upon both of these properties.

Question 195

T or F - Concerning MRI in the context of imaging the heart:
A. Magnetic resonance imaging does not use ionising radiation
B. Motion of the heart will not affect the quality of MR images obtained using standard imaging techniques
C. Magnetic resonance imaging can acquire tomographic images at any orientation in the heart

Answer 195

A. True. Magnetic resonance imaging is a tomographic technique that does not use ionising radiation.

B. False. Most conventional MRI techniques take several minutes to acquire, during which time significant cardiac and respiratory motion could occur, leading to artefacts in the images. Even very fast techniques that can be performed in a single breath hold may take up to 20 seconds (s) to acquire. In order to image the heart using MRI, special techniques must be used that can effectively ‘freeze’ the motion of the heart and the chest.

C. True. It is possible, using MRI, to acquire tomographic images of the heart in any orientation.

Question 196

How do you eliminate movement artefact in cardiac MRI?

Answer 196

synchronise the acquisition to the cardiac + respiratory cycle/breath-holding

Question 197

What is the magneto-hydrodynamic effect?

Answer 197

When an ECG is being acquired for physiological monitoring, the leads are placed far apart to maximise the electric potentials. If the patient is in an MRI scanner, this can lead to problems, as blood flowing through the magnetic field creates electric fields that are detected as additional voltages by the ECG electrodes

Question 198

Why is the magneto-hydrodynamic effect an issue?

Answer 198

It is the blood flow in the aorta during systole that generates the greatest voltages. These superimpose on the T-wave, elevating it to the point where it may be greater than the height of the R-wave. This may interfere with the way the scanner synchronises data acquisition to the cardiac cycle.

Question 199

How can you mitigate for the magneto-hydrodynamic effect?

Answer 199

place the ECG leads closer together

vector ECG gating, which takes into account the direction of the electric fields as well as the magnitude.

peripheral pulse (PP) gating

Question 200

How do you acquire an ECG gated MRI?

Answer 200

all the raw data must be obtained from the same point in the cardiac cycle. This is done by using the R-wave of the ECG (or the peak of the PP signal) as a reference point for the scanner. The scanner can automatically monitor the ECG and begin scanning when it detects the R-wave. A variable trigger delay can be set so that data can be obtained from any point in the cardiac cycle.
The image is built up over n heartbeats, where n is the number of phase-encoding steps. The repetition time (TR) of the sequence is fixed by the R-R interval.

Question 201

how can you get moving images of the heart with MRI?

Answer 201

this is not a ‘real-time’ display as data for the individual images are built up over several cardiac cycles in a similar way to ordinary cardiac gating. In order to be able to do this within a breath-hold, very fast MRI scans must be used. The two techniques most commonly applied are the T1-weighted (T1W) fast spoiled gradient-echo (GE) or steady state free precession (SSFP) GE. These sequences have TR values of the order of a few ms.

Question 202

What is segmented cine imaging and why is it used?

Answer 202

Collecting only one line of k-space per image per R-R interval means that the scan time will be the R-R interval times the number of phase-encoding steps. For a typical image containing somewhere between 128 and 256 phase-encoding steps, the scan time will be excessively long, with motion artefacts arising from breathing obscuring the heart. The solution is to use segmented cine imaging.

Segmented cine imaging acquires more than one line of k-space per image in each R-R interval. This means that scans can be performed within a single breath-hold.

Question 203

What is the difference between prospective and retrospective gating?

Answer 203

In prospective gating (or triggering), data acquisition begins only after the detection of a physiological event (such as the ECG R-wave).

In retrospective gating, data acquisition is continuous and is not initiated by a cardiovascular trigger event. After the end of acquisition, the MR data may be reordered, grouped or correlated with the phase of the cardiac cycle.

Question 204

What type of gating is normally used for cardiac MRI?

Answer 204

retrospective gating

Question 205

How does retrospective gating work?

Answer 205

data are acquired continuously throughout the cardiac cycle, with the R-wave causing a real-time update of the phase-encoding gradient. The ECG is recorded simultaneously with the MR data for the whole scan time. The data are sorted after acquisition to account for variations in the R-R interval. Each cardiac cycle is divided into the required number N of cardiac phases. This is done by measuring the duration of each R-R interval and calculating the time of each temporal cardiac phase as a percentage of that interval. This provides a linear expansion or contraction of each cardiac cycle to account for variations in the duration of individual heartbeats. The nearest acquired phase-encoding step to each of the N temporal phases is placed in the raw data matrix for the image at that phase. Once all the temporal phases have been filled with their nearest phase-encoding steps, the reconstruction proceeds as normal. Each acquired phase-encoding step may be used for more than one of the N cardiac phases.

Question 206

How does prospective gating work?

Answer 206

In prospective gating, the mean R-R interval is measured over a number of cardiac cycles.

Following this measurement, acquisition of MR data is triggered by the R-wave of the ECG (even though the pulse sequence may run continuously). Once triggered, acquisition is repeated continually for 85-90% of the mean R-R interval with one line of k-space being added for each cardiac phase image for a particular slice. The remainder of k-space is filled over many subsequent cardiac cycles.

Depending on the heart rate, the number of cardiac phases required and the value of TR, one or two slice locations may be acquired.

At the end of each cardiac cycle, an arrythmia rejection (AR) window is set to allow for variations in heart rate; typically, its duration is 10-15% of the mean R-R interval. If the next R-wave occurs outside the AR window (i.e. much earlier or much later than expected), all the associated data are discarded and re-acquired. Furthermore, data acquired at the end of the cardiac cycle during the AR period are also discarded, even if the next R-wave occurs within the window; this gives a ‘dead-time’ at the end of each cycle.

Question 207

What artefact can you get with prospective gating?

Answer 207

There is T1 recovery of the magnetisation during the AR period resulting in a bright first image following the R-wave; this gives a flashing artefact when the images are replayed as a cine loop.

Question 208

How is ventricular function calculated on cardiac MRI?

Answer 208

series of Short Axis images is taken of the heart, from base to apex. Drawing contours around the endocardial and epicardial borders allows the myocardial mass and blood pool volumes to be calculated.

Question 209

What are the normal cardiac MRI sequences known as?

Answer 209

‘white blood’ imaging, due to the high signal intensity of the blood pool.

Question 210

What is black blood cardiac MRI?

Answer 210

Cardiac gated fast spin echo (FSE) sequences with T1W produce a single high-resolution image of the cardiac anatomy. However, they can suffer from artefacts due to the inflow of blood into and out of the imaging slice. A special technique called black blood imaging is used to reduce these artefacts.

Question 211

How are black blood cardiac MRI images acquired?

Answer 211

double inversion recovery (DIR) FSE. This means the blood in the imaging slice should flow out of the slice during the TI period and be replaced with inverted blood from outside the slice.

Question 212

How is cardiac perfusion MRI performed?

Answer 212

The technique uses an injection of gadolinium contrast agent and then repeatedly scans the heart as the contrast arrives and leaves. The contrast agent will enhance myocardium on T1W sequences where the perfusion is good. Poor perfusion may only be evident under conditions where the heart is made to work hard, known as stressing. This is usually achieved using pharmacological agents such as dobutamine or adenosine. The agent is infused intravenously over a period of several minutes prior to scanning.

Question 213

What is delayed enhancement Cardiac MRI imaging?

Answer 213

allows an idea of whether re-perfusion will be successful.

myocardium will exhibit a rapid increase in signal on T1W sequences as the contrast arrives, and then a steady wash out over the next five to ten minutes. Infarcted but viable myocardium demonstrates reduced enhancement at a lower rate, but there is still wash out of the contrast. In non-viable myocardium, cellular breakdown allows the contrast to accumulate within the tissue without being washed out.

Question 214

Regarding motion artefacts in cardiac MRI, which of the following statements is/are accurate?

A. They may be caused by both respiratory and cardiac motion
B. They mainly occur in the frequency encoding direction
C. They do not affect the quality of diagnostic images
D. They mainly occur in the phase encoding direction
E. They appear as ‘ghosts’

Answer 214

A. Correct. The movement of any structure during image acquisition will cause motion artefacts.

B. Incorrect. Each row of k-space data in the frequency encoding direction is usually quickly filled, and so significant motion is unlikely during this period.

C. Incorrect. Motion artefacts degrade image quality.

D. Correct. Each column of k-space data in the phase encoding direction is filled in the total image acquisition time, during which significant motion may occur.

E. Correct. The ‘ghosts’ are those of the edges of moving structures.

Question 215

In cine cardiac MRI, the temporal resolution achieved with 20 phases in a patient with a normal heart rate is about…?

Answer 215

The normal heart rate is about 60 beats per minute, which means that the R-R interval is about 1 s or 1000 ms. With 20 phases, the temporal resolution is 1000/20 = 50 ms.

Question 216

What are the 4 main MRI contrast types in liver imaging?

Answer 216

Dynamic phase T1 paramagnetic (extracellular) contrast agents

Delayed phase T1 hepatobiliary contrast agents

Combined T1 extracellular and hepatobiliary contrast agents

T2 contrast agents that target the reticuloendothelial system

Question 217

How do paramagnetic contrast agents work?

Answer 217

by causing relatively large local magnetic field distortions due to their large magnetic moments. These enhance the T1 and T2 relaxation of protons that come into close proximity.

When placed in a strong magnetic field the contrast agent develops a large magnetic moment, which enhances the relaxation rates of protons in any water molecules that approach the vicinity of the agent. In fact, although it is referred to as a T1 contrast agent, the effect is such that both T1 and T2 relaxation times are reduced. However, at standard clinical doses the T1 relaxation effect dominates.

Question 218

What is a commonly used paramagnetic contrast agent?

Answer 218

One of the primary constituents of an MRI paramagnetic contrast agent is the metal ion gadolinium (Gd).

Question 219

What factors effect the degree of T1 relaxation with paramagnetic contrast agents?

Answer 219

Concentration of the agent (Gd(read a full definition of this term)) in the tissue
Proximity (D) of the agent to tissue protons that are undergoing relaxation
Rotational motion (tumbling) (R) of the agent
Number (C) of water molecules that associate with the agent
Time (T) that the water molecules are available for association with the agent

If some or all of these conditions are met then T1 relaxation will be optimised and this will result in increased signal intensity on T1W images.

Question 220

describe the response graph for gadolinium contrast.

Answer 220

there is a steady increase in signal intensity as the concentration of the agent is increased, until an optimum concentration (giving maximum intensity) is reached.

Once the optimal concentration is exceeded then the effect of introducing more contrast agent actually causes the signal intensity to decrease. At lower concentrations, T1 relaxation effects dominate, but at higher concentrations T2 relaxation effects begin to dominate.

Question 221

Does the signal intensity of the tissues change characteristically over time with gadolinium contrast?

Answer 221

yes. It is possible to use the dynamic information available from the contrast agent signal intensity changes over time to characterise the various types of lesion that might be present. In other words, contrast enhanced MRI studies offer not just improved anatomical information, but also additional dynamic information related to the specific contrast uptake activity of the target tissues. In order to obtain these dynamic images, the contrast agent is injected as a bolus and T1W images are acquired at various times (phases) after the injection.

Question 222

What are the normal contrast phases for gadolinium contrast?

Answer 222

The arterial phase (typically 30 seconds post injection)
The venous phase (typically 60-90 seconds post injection)
The equilibrium phase (typically 3-5 minutes post injection)

Question 223

How do delayed phase hepatobilliary contrast agents work?

Answer 223

taken up by functioning hepatocytes in healthy liver tissue. The agent causes a reduction in the functioning liver tissue T1, resulting in hyperintense healthy liver tissue relative to hypointense lesion regions on T1W images.

The T1 shortening is achieved by the presence of manganese (Mn) in the agent, which conveys paramagnetic properties. These agents can also be used for imaging 24 hours post-contrast since the manganese gradually accumulates within any non-healthy liver tissue during this time course. The contrast agent takes longer to be released from such lesions than it does from healthy liver tissue. Therefore, at 24 hours, the presence of hepatobiliary contrast agent in the lesion causes T1 shortening of the lesion and therefore hyperintense signal intensity on a T1W sequence relative to the more hypointense healthy liver tissue.

Question 224

What is the mode of action for T2 contrast agents?

Answer 224

the presence of the contrast agent causes the precessional frequency of the water proton to temporarily speed up or slow down. As soon as the water proton moves away from the contrast agent then it will continue to precess at its original frequency, but it will have gained or lost phase as a result of its brief encounter with the contrast agent field.

This dephasing is accelerated in the presence of a contrast agent

Question 225

Describe the typical response curve for a T2 contrast agent.

Answer 225

There is an exponential decrease in signal intensity (negative enhancement) as the concentration of the agent is increased.

In other words, as the concentration becomes larger, signal dephasing becomes more prominent, and the signal intensity on a T2W image becomes lower.

Question 226

What is the effect of MRI contrast agents on the tissue relaxation times T1 and T2?

A. A T1 agent reduces T1 and has no effect on T2
B. A T2 agent increases T2 and reduces T1
C. A T2 agent reduces T2 and has no effect on T1
D. A T1 agent reduces T1 and increases T2
E. All agents reduce both T1 and T2

Answer 226

The correct answer is E.

The magnetic fields associated with all types of MRI contrast agent serve to reduce both T1 and T2. Which effect is used to change the contrast between tissues depends on what type of image weighting is chosen; image weighting is determined by pulse sequence. For example, a post-contrast T1W image uses the reduction in T1 to change contrast and a material that is used in this way is called a T1 contrast agent. The contrast between a lesion and surrounding healthy tissue may be due to relatively greater concentration of the agent in either; this depends on factors such as blood flow and physiological uptake.

Question 227

What TE and TR gives T1 contrast?

Answer 227

Short TE
Short TR

Question 228

What TE and TR gives T2 contrast?

Answer 228

Long TE
Long TR

Question 229

What TE and TR gives proton-density contrast?

Answer 229

Short TE
Long TR

Question 230

What is Slice separation in MRI?

Answer 230

The distance between adjacent slices is referred to as the slice separation. Often, the gap in between adjacent slices (i.e. the slice gap) is quoted. Slices with no slice gap are called contiguous, as they cover the entire scanning region

Question 231

What is crosstalk?

Answer 231

Acquiring contiguous images may result in crosstalk, where the MRI signal from one slice interferes with the signal from another.

Question 232

Why does crosstalk occur and what effect does this have on the image?

Answer 232

Crosstalk occurs because the slices are not perfectly rectangular and overlap. Contrast in the final image may be altered as magnetisation in the overlap regions experiences more RF pulses per TR interval.

Question 233

What is the only way to avoid crosstalk?

Answer 233

The only way to avoid crosstalk is to increase the gap between the slices (in which case they are no longer contiguous) or change the order in which the slices are acquired (and maintain contiguity).

Slices can be acquired in one of two ways:

Sequential
Interleaved

Question 234

What is sequential multislice imaging?

Answer 234

Sequential multislice imaging acquires all the image data for one slice before moving onto the next

Question 235

What is the advantage of sequential multislice imaging?

Answer 235

The advantage of this technique is that crosstalk is avoided, since only one slice is acquired at any one time. This applies both to a set of contiguous slices and a set with a finite slice gap.

Question 236

What is the disadvantage of sequential multislice imaging?

Answer 236

scan times can be very long as the time taken to acquire one image is now multiplied by the number of slices. This technique is suitable if the sequence TR (and therefore the time required to acquire a single slice) is very short.

Question 237

What is interleaved multislice imaging?

Answer 237

For sequences where the TR is long, there will be a significant time between the end of data acquisition and the beginning of the next RF pulse. All that happens during this time is relaxation. This time can be used to acquire the image data for all the other slices, reducing the total acquisition time to the time required to acquire just one slice. This is called interleaved scanning.

This works because RF pulses only affect protons that have the correct Larmor frequency.

Question 238

What is the disadvantage of interleaved multislice imaging and how can this be mitigated?

Answer 238

here is the possibility of crosstalk occurring if the slices are contiguous or close together, as the time between acquiring data for adjacent slices is very small. This can be made larger by altering the order of slice acquisition, or by concatenating the scan , where adjacent slices are not acquired together. Concatenating the scan can also be useful in reducing the TR.

Question 239

What is the benefit of 3D MRI imaging over 2D MRI imaging?

Answer 239

3D scan, where the data is acquired as a volume rather than a series of individual slices. The data can then be divided into slices after reconstruction. 3D scans:

Produce perfectly contiguous slices with no crosstalk

Tend to produce thinner slices than 2D sequences

Question 240

What does 3D imaging allow you to do?

Answer 240

view 3D images at different orientations and even generate surfaces and calculate volumes. Images at other orientations are called multiplanar reconstructions (MPRs). From a single set of 3D data, it is possible to display images in sagittal, transverse and coronal orientations. Rendered images show the surface, and it is possible to cut out parts of the volume to view the underlying anatomy.

Question 241

How is 3D MRI acquired?

Answer 241

A thick slab is excited by the RF pulse. This is achieved by using very short pulse durations and/or a low gradient strength.

Spatial localisation in the slice-select direction is achieved by phase encoding in two directions, rather than just one. The two phase encodings do not interfere with each other as they are generated by gradients that are at right angles. The data is stored in a 3D array and reconstructed using a 3D Fourier transform.

Question 242

What is a fast spin/turbo spin echo sequence?

Answer 242

multiple echoes can be used to obtain extra phase-encoding steps. This can reduce the scan time dramatically as more lines of data are acquired in a single TR period. This technique is commonly used to shorten the acquisition time of T2-weighted (T2W) SE images that have a long TR.

Question 243

What is the echo train length?

Answer 243

The number of echoes acquired for each excitation (90°) RF pulse is called the echo train length (ETL).

Question 244

How does the ETL effect scan time?

Answer 244

The scan time will be shortened by a factor given by the ETL. For example, if the scan time of an SE sequence were 4 minutes and 16 seconds, using an FSE with an ETL of 4 (and all other parameters the same) would shorten the time to 1 minute and 4 seconds.

Question 245

What is the “effective TE” of an MRI scan?

Answer 245

The TE of an FSE sequence is often defined as being the TE of the echo that corresponds to the centre of k-space (i.e. the echo with no phase-encoding gradient, where the echo amplitude is maximum). This is referred to as the effective TE (TEeff). The images suffer from a loss in resolution in the phase-encoding direction, as the echoes are not all acquired at the same TE, and there is significant T2 decay between subsequent echoes. This loss of resolution, so-called ‘T2 blurring’, gets worse when high ETLs are used.

Question 246

What is a single shot FSE?

Answer 246

Echo train lengths for FSE sequences can be increased to the point where all the data required to create the image are obtained after a single RF excitation. For example, a 256 x 256 image would require an ETL of 256. However, a technique called half-Fourier imaging is used to reduce this to just over half that number. An entire image can be obtained in about 2 seconds, allowing regions such as the abdomen to be scanned without motion artefacts.

Question 247

What is echo-planar imaging?

Answer 247

another technique that allows extremely rapid data acquisition.

As with the SSFSE, all the data is acquired after a single RF excitation. However, EPI does not use trains of 180° pulses nor does it reverse each phase-encoding gradient before acquiring the next line of data. the phase-encoding gradient is ‘blipped’ every time the frequency-encoding gradient is switched, so that each line of data will have a different accumulated phase. the raw data is built up in one go by following an ‘S’ shape through k-space.

Question 248

What is EPI factor and why is it important for EPI imaging?

Answer 248

the EPI factor determines how many lines of data are collected for each image. Using a high EPI factor will result in a high-resolution image; however, the minimum achievable TE may be quite high as well, which will affect the contrast in the image.

Question 249

When can a lower flip angle be better for MRI?

Answer 249

using a lower flip angle, for example 30 ° results in completely different curves than 90.

In the case of SE sequences, the signal is lower for all TRs; therefore there is no benefit to using a reduced flip angle.

The signal from a low flip-angle GE sequence is actually greater for short TRs. Therefore, there is a benefit in using a short TR, low flip angle GE sequence for fast imaging.

Question 250

What needs to be added to short TR GE images?

Answer 250

Fast GE sequences are similar to normal GE sequences except that they contain additional components that prevent the formation of artefacts, due to the short values of TR used. Extra gradients known as spoilers are added, which prevent any residual transverse magnetisation from interfering with subsequent echoes.

Another way of removing this residual transverse magnetisation is to use RF spoiling, where the phase of the RF pulses is changed.

Question 251

What is the Ernst angle?

Answer 251

optimum angle that gives a maximum steady state MRI signal. This depends on the T1 of the tissue and the TR of the sequence.

Question 252

What is a hahn echo?

Answer 252

A partial SE generated from two previous RF pulses, and a stimulated echo from three RF pulses.

Question 253

What is a coherent steady state?

Answer 253

it is possible to generate signals that do have some T2 dependence using combinations of RF pulses, if the transverse magnetisation is not spoiled at the end of each TR. A partial SE (called a Hahn echo) can be generated from two previous RF pulses, and a stimulated echo from three RF pulses. The total MRI signal will be the sum of all these contributions. This is called a coherent steady state.

Question 254

What sequence uses coherent steady state?

Answer 254

steady state free procession

For every positive gradient, there is a negative gradient that prevents transverse magnetisation from being spoiled, allowing the build-up of a coherent steady state.

These sequences are run with very short values of TR and TE. This is done to reduce the dephasing effects of magnetic field inhomogeneities (so called resonance-offset effects). These can cause the various components of the coherent steady state to interfere destructively, producing dark bands across the images.

Question 255

What does SSFP contrast depend on?

Answer 255

depends upon the ratio of T2 to T1. This is only true for large flip angles. This means that tissues where T2 is similar to T1 (for example, water and cerebrospinal fluid (CSF)) will give the highest signal (remember that T2 is always shorter than T1).

Question 256

Regarding SE imaging:
A. Crosstalk cannot occur with the interleaved acquisition of contiguous slices
B. Three-dimensional imaging can produce contiguous slices with no crosstalk
C. A thick slab of tissue is excited with a long RF pulse in 3D imaging
D. In multiple contrast imaging, the phase encoding gradient is changed between each echo
E. Half-Fourier imaging increases the acquisition time in SSFSE

Answer 256

A. Incorrect. With interleaved multislice imaging, crosstalk is very likely to occur if a set of contiguous slices is acquired in numerical order. Crosstalk may be avoided by changing the order of slice acquisition or by concatenating the scan.

B. Correct. With 3D acquisition, data are obtained from a block of tissue and may be reconstructed as a set of thin contiguous slices with no crosstalk.

C. Incorrect. In 3D imaging, a short RF pulse is used to excite a thick slab of tissue because the pulse may be represented by a wide range of frequencies.

D. Incorrect. In multiple contrast imaging, for each phase encoding gradient a succession of 180° refocusing pulses are used to generate a series of echoes with different values of TE.

E. Incorrect. In SSFSE, half-Fourier imaging reduces image acquisition time.

Question 257

Regarding GE imaging:

A. It is the only type of sequence for which EPI can be performed
B. For fast imaging, a short TR and a large flip angle are used
C. Spoiler gradients produce artefacts in fast imaging
D. Setting the flip angle to be equal to the Ernst angle gives the minimum steady state MRI signal
E. SSFP pulse sequences have very short values of TR and TE

Answer 257

A. Incorrect. Echo-planar imaging can be done with both GE and SE pulse sequences.

B. Incorrect. A short TR and a small flip angle are used for fast imaging.

C. Incorrect. Spoiler gradients reduce artefacts by preventing residual transverse magnetisation from interfering with subsequent echoes.

D. Incorrect. Using the Ernst angle gives the maximum steady state signal for a particular tissue.

E. Correct. As well as giving very short image acquisition times, the use of short TR and TE reduce the dephasing effects of magnetic field inhomogeneities.

Question 258

What is the basic principle of time-of-flight MRA?

Answer 258

The principle of TOF-MRA is to use the motion characteristics of flowing blood to generate contrast between blood and surrounding tissue

Question 259

Why is a transverse slice for TOF-MRA normally used?

Answer 259

It is important to understand that the imaging slice needs to be placed roughly perpendicular to the axis of the artery through which blood is flowing

Question 260

How does TOF-MRA differ between SE and GE?

Answer 260

During the time taken between the application of a radiofrequency (RF) pulse and the acquisition of the MRI signal (the TE time), movement of blood through the imaging slice results in very significant contrast changes. In a SE sequence, a high blood velocity causes blood within the imaging slice to leave it between the slice selective 90° excitation and 180° refocusing pulses. Thus, no echo is generated and there is no signal from the vessel lumen giving the appearance of ‘black blood’.

In a GE pulse sequence, an RF pulse is applied, followed by phase encoding and echo signal detection during frequency encoding – all completed within the sequence repetition time (TR). The gradient used to create the echo is not slice selective. This means that blood within the slice that receives an RF pulse produces an echo even though it has left the slice. Furthermore, by the time the next RF pulse is applied, this ‘saturated’ blood has been replaced within the slice by fresh ‘unsaturated’ blood (with full longitudinal magnetisation) and so the vessel lumen generates high signal compared with the surrounding static tissue giving the appearance of ‘bright blood’.

Question 261

What sequence is generally used for TOF-MRA?

Answer 261

bright blood using a GE pulse sequence

Question 262

Why does blood look bright on GE TOF-MRA?

Answer 262

the arterial blood is not stationary, and moves through the slice at a velocity that is dependent on the area of anatomy under investigation. Again, if an RF excitation pulse is applied, then the blood magnetisation and the surrounding tissue magnetisation are both excited into the transverse plane.

However, during the TR time of the sequence, the blood originally in the slice, flows out of the slice and is replaced by new ‘fresh’ blood. In other words, the new fresh blood has managed to enter into the imaging slice without experiencing the initial RF pulse of the sequence. This new blood is already at its equilibrium magnetisation and effectively looks as though it has an infinitely short T1 relaxation time.

On a T1-weighted (T1W) gradient echo (GE) sequence therefore, flowing blood is clearly visible as hyperintense signal relative to the surrounding tissues. This phenomenon is often referred to as inflow enhancement

Question 263

When is inflow enhancement greatest?

Answer 263

if TR ≥ dZ/v where dZ is the slice thickness and v is the blood velocity.

Question 264

What factors affect contrast in TOF-MRA?

Answer 264

Blood velocity
Direction of blood flow relative to slice plane
Slice thickness and slice profile (type of RF pulse)
Sequence flip angle
Sequence TR
T1 relaxation times of the tissues
Use of saturation bands

Question 265

When does maximal contrast occur in TOF-MRA?

Answer 265

when the velocity of the blood is greater than the imaging slice thickness divided by the TR time

Question 266

What is the advantage and disadvantage of 2d TOF-MRA?

Answer 266

Using a sequence with a short TR and a large flip angle will generate excellent contrast between the (hyperintense) vessels and the background (hypointense) stationary tissues. The downside of the 2D TOF technique, however, is that the slice resolution is actually fairly coarse and the slices are not quite contiguous.

Question 267

What is the advantage and disadvantage of 3D TOF-MRA?

Answer 267

The advantage of 3D TOF-MRA , is that it offers better ‘slice’ resolution than 2D TOF-MRA, thus allowing isotropic voxel acquisitions. However, the contrast between the vessels and the background tissue signal is generally poorer than for 2D TOF-MRA techniques.

Question 268

How are MIP images formed in TOF-MRA?

Answer 268

Maximum intensity projection images are formed by observation of maximal pixel values at incremental planes around the imaging volume

Question 269

Why is the contrast worse in 3D TOF-MRA than 2D?

Answer 269

In 3D TOF, a particular region of flowing blood experiences a number of different RF pulses as it flows across the slice block. This in turn will alter the level of saturation of the blood across the laminar flow profile of the vessel as a function of the depth of the imaging block.

As blood flows through the slice block it becomes progressively more saturated by the application of repeated RF pulses. The faster blood (central in the vessel) is less prone to this effect since it is flowing faster and therefore experiences less RF excitation across the slice block. Conversely, the slower blood (at the edges of the vessel) experiences more RF excitation, and therefore becomes more saturated, resulting in less contrast relative to the stationary tissue surrounding the vessel lumen. The effect on the images, is a progressively smaller observed vessel lumen as the distance into the slice block increases.

Question 270

How do you solve issues with contrast in 3D TOF-MRA?

Answer 270

1. Tilted optimal non-saturating excitation pulse
   The principle of TONE pulse is that the flowing blood experiences a progressively larger flip-angle as it moves through the block of tissue that is imaged. This helps to ensure that even saturation is maintained across the block.
2. Multiple overlapping thin-slab acquisition
   essentially a hybrid of 2D and 3D TOF. The idea behind the technique, is that multiple slabs along the vessels are acquired, but each slab is actually a 3D volume.

The slabs overlap to ensure full coverage of the vessel, and the slab thickness (dZ) is much smaller, which means that the blood remains unsaturated as it crosses the slab and the resulting contrast between the high signal blood and the low signal stationary tissue is maintained.

Question 271

Why do you need to use a long TR in 3D TOF-MRA and why can this be an issue??

Answer 271

in order to avoid significant saturation of the flowing blood through the slice block.

An obvious downside of this, is that suppression of the stationary tissues surrounding the vessels is less effective.

Question 272

How do you solve problems with long TR in 3D TOF-MRA?

Answer 272

One solution is to implement magnetisation transfer contrast (MTC) in order to suppress the signal from these stationary tissues.

Magnetisation transfer is a technique that relies on the two-way exchange of magnetisation between protons in different molecular environments within the tissue. In simple terms, protons can exist in a ‘free’ state, for example in blood or extracellular fluid or in a ‘bound’ state (associated with larger macromolecules). Magnetisation transfer (two-way exchange) will occur in tissues where ‘free’ and ‘bound’ water protons co-exist in close proximity. Blood has very little ‘bound’ water component and is therefore not susceptible to MTC signal changes, but other tissues such as grey matter and white matter are far more susceptible to MTC changes.

Magnetisation transfer is induced by the application of off-resonance RF pulses that affect only the ‘bound’ water protons. The off-resonance RF causes saturation of the ‘bound’ water protons, and this results in transfer of magnetisation from the free water state to the bound water state. In other words, the exchange has changed from a two-way process into a one-way process.

The final images are acquired a short time after the MTC pulses have been applied, and tissues that are MTC susceptible will demonstrate signal loss relative to those that are not MTC susceptible.

Question 273

How is signal from veins removed from TOF-MRA?

Answer 273

by applying RF pulses in the form of a ‘saturation band’ in a region that is manually placed above the plane of the imaging slices.

Using this technique, the venous blood that is flowing (returning) into the slice, will be saturated prior to the imaging sequence, and so any venous blood that flows into the plane of the imaging slices during the TOF process, will not contribute any signal to the final image.

Question 274

How is black blood imaging performed in TOF-MRA?

Answer 274

the same basic properties of the TOF-MRA technique, but with the addition of inversion pulses to run the sequence as an inversion recovery (IR) technique.

Apply two 180° inversion pulses; the first one is applied to the entire tissue sample, and the second one is slice selective. all protons outside the imaging slice to experience a 180° pulse, whilst those protons inside the imaging slice have experienced both 180° pulses (i.e. the equivalent of a 360° pulse).

a HASTE or a turbo spin echo (TSE) is applied following an inversion time (TI) of approximately 600 ms (at 1.5 T), then the blood that was originally outside the imaging slice will have moved into the slice at exactly the time when its magnetisation is returning through the null point for blood.

Question 275

What is phase contrast MRA?

Answer 275

Using this technique, it is possible to generate velocity encoded images related to the flow rate of the blood. Bipolar gradients are applied in order to induce a measurable phase change that is related to the strength of the applied gradient and the blood flow velocity.

the effect of the equal and opposite bipolar gradients is to induce phase changes that add to zero for stationary protons, but result in a net phase shift for moving protons such as blood. The amount of phase change (θ) is related to the blood flow velocity.

Question 276

What is the most common method for MRA?

Answer 276

Contrast enhanced MRA

Question 277

How does contrast-enhanced MRA work?

Answer 277

The process involves imaging the vessels of interest before contrast agent delivery, and then again following delivery of a bolus of (usually) a gadolinium-based extracellular contrast agent. The contrast agent causes a large reduction in the blood T1 from approximately 1200 ms to around 50 ms (at 1.5 T).

The timing of the data acquisition is critical. Acquire the data too early and the contrast agent will not have reached the vessels of interest, but acquire the data too late and the contrast will have reached the venous system. With careful timing, it is possible to obtain a subtraction image of pre-contrast minus post-contrast datasets. The subtraction technique will effectively remove any MRI signal associated with non-vascular regions, leaving just the contrast-enhanced vessels visible.

Question 278

What sort of sequences are used for CE-MRA?

Answer 278

The most commonly used sequence for CE-MRA is a fat-suppressed 3D spoiled fast gradient echo sequence (often used with parallel imaging). As previously discussed, the sequences used are T1-weighted in order to take advantage of the fact that the contrast agent bolus causes T1 shortening of the blood.

Question 279

What are the potential pitfalls with CE-MRA (4)?

Answer 279

Turbulent flow
In the presence of a stenosis, the blood flow immediately after the occluded region often becomes turbulent. The presence of turbulent flow can lead to phase related signal void in the region and therefore make the occlusion appear rather larger than it actually is.

Tortuous vessels
The shape of the vessel may also cause a problem with over-estimation of stenotic regions. This will result in exaggerated regions of signal loss, more due to the tortuous plane of the vessel rather than the presence of a stenosis.

The choice of imaging slice thickness is also very important. If the slice is too thick, then the contrast between the vessel and the stenosis will deteriorate to the point where the images are clinically non-diagnostic. Conversely, if the slice thickness is too thin, then it is possible that the slice will not actually encompass the stenotic areas, and the narrowing may only be partially covered, or even missed.

If the imaging matrix is too coarse then small subtle stenotic areas may be missed due to the signal averaging that will occur within each pixel. The goal is to acquire the images at the highest resolution achievable, provided that the data can be acquired comfortably within a breath hold and the SNR is not degraded too significantly.

K space encoding time
Conventional data collection is achieved by sampling each line of data sequentially in the phase encoding direction, usually from the centre. This can pose a problem in CE-MRA, since the time taken to acquire all lines of data can take up to 30 seconds, which might result in images with venous contamination. Since it is known from earlier sessions that the central region of k-space contains the image contrast information and the outer region of k-space contains the image detail, e.g. boundary information, then it follows that optimal arterial phase contrast will be obtained if the central regions are acquired quickly in the first instance.

Question 280

WHat is a risk with gadolinium contrast?

Answer 280

administering gadolinium-based contrast agents to patients with renal failure are more prone to developing nephrogenic systemic fibrosis (NSF)

Question 281

If the inflowing blood velocity is greater than the slice thickness divided by the sequence TR, how will the blood appear on a T1W TOF GE sequence?

Answer 281

hyperintense
The reason for this is that the unsaturated blood moving towards the slice is flowing fast enough to completely replace the saturated blood within the slice.

Question 282

What combination of resolution parameters is most appropriate for CE-MRA?

A. Thin slice, high in-plane resolution
B. Thick slice, high in-plane resolution
C. Thin slice, low in-plane resolution
D. Thick slice, low in-plane resolution

Answer 282

The correct answer is A.

However, care must be taken to ensure that the source images are carefully reviewed to ensure that stenoses are not accidentally ‘missed’ by the thin slices. Also, the SNR should be maintained at a high enough level such that image quality is not significantly degraded.

Question 283

What are the main factors that influence diffusion?

Answer 283

The main parameters that influence diffusion are the temperature of the material or substance (proportional to the kinetic energy), the mass of the diffusing particles, and the viscosity of the environment (e.g. gas, liquid or solid). Diffusion is more rapid in a gas, less rapid in a liquid and much slower in a solid, due to the way in which the constituent atoms are bonded in each case.

Question 284

What does diffusion weighted MRI allow imaging of?

Answer 284

Diffusion-weighted MRI (DW-MRI or simply DWI) allows imaging of apparent diffusion coefficient (ADC). This may be calculated with knowledge of the degree of diffusion weighting (b-value), which depends on the characteristics of special diffusion gradients used in the MRI pulse sequence.

Question 285

Where is perfusion MRI used?

Answer 285

perfusion MRI is often applied to the brain, in which case it is used to image quantities such as relative cerebral blood flow (rCBF) and relative cerebral blood volume (rCBV). However, it may also be applied to tissue in other organs such as kidney and heart. The two main techniques are dynamic contrast-enhanced MRI (DCE-MRI) and arterial spin labelling (ASL).

Question 286

What is brownian motion?

Answer 286

the mean square displacement (in one direction) of particles undergoing Brownian motion is given by: 2Dt where D is the diffusion coefficient and t is the elapsed time

Question 287

What is the diffusion pulse sequence?

Answer 287

the pulsed gradient spin echo (PGSE) sequence. It is essentially a spin-echo sequence with two additional large symmetrical gradients (diffusion gradients) placed either side of a 180° refocusing pulse.

In diffusion MRI, the duration of each diffusion gradient pulse is usually labelled δ, and the time between the onset of the first diffusion pulse and the second is usually labelled Δ. The strength of each diffusion gradient is usually labelled G

Question 288

how does the diffusion sequence effect stationary protons?

Answer 288

the effect of the first diffusion gradient is to induce spin dephasing in the tissue. Application of a 180° pulse provides the condition for the transverse magnetisation to commence rephasing. The application of the second diffusion gradient then causes the transverse magnetisation to return into phase. On stationary protons, the effect of the second diffusion gradient acts to rephase the dephasing caused by the first diffusion gradient. The net effect for stationary protons is that no dephasing, or signal loss, occurs.

Question 289

how does the diffusion sequence effect diffusing protons?

Answer 289

The situation is different for moving (diffusing) protons

the initial dephasing of the transverse magnetisation caused by the first diffusion gradient is exacerbated by additional phase changes due to molecular diffusion.

Consequently, the actual phase changes will be greater than those due to the application of the diffusion gradient alone. After the application of the 180° pulse and the second diffusion gradient, further molecular diffusion has occurred such that the application of the second gradient is not able to fully ‘undo’ the effect of the first gradient The net result is a degree of transverse spin dephasing, and hence signal loss, which is proportional to the amount of molecular diffusion that has taken place.

Question 290

What is the b-value in DWI?

Answer 290

The b-value is a numerical representation of the degree of diffusion weighting imparted by the diffusion gradients during the MRI sequence. The mathematical derivation is not included within this course, but the final expression is as follows: b=(YGδ )^2 Δ

where the quantities G, δ and Δ refer to the strength, duration and temporal separation of the diffusion gradients and Y the gyromagnetic ratio (constant)

The higher the b-value the stronger the degree of diffusion weighting.
the units of b are s mm^-2.

Question 291

What is ADC?

Answer 291

In MRI, D is called the apparent diffusion coefficient (ADC). D is usually expressed in units of mm2 s-1. A quantitative image of D may be calculated on a pixel-by-pixel basis and this parametric image is known as an ADC map

Question 292

What is T2 shine through?

Answer 292

In areas of damaged brain tissue, for example following stroke, the sudden development of tissue ischaemia often results in corresponding areas of brain with a significantly reduced diffusion coefficient. On a diffusion-weighted image therefore, the amount of diffusion-related signal loss that would normally occur in healthy brain tissue is lowered, and the lesion therefore appears hyperintense. However, since diffusion-weighted sequences tend to be inherently T2-weighted (T2W), the image is also likely to contain areas of hyperintense signal on diffusion-weighting imaging (DWI). This obviously raises the possibility of false-positive diagnosis

Question 293

How can T2 shine-through be resolved?

Answer 293

Either observe qualitatively what happens to each lesion when diffusion weighting with larger b-values is used, or
Generate an ADC map. If an ADC map is constructed, then it is possible to distinguish each lesion by virtue of the ADC - in other words the T2-weighting component is diminished

Question 294

What are the typical signal changes in stroke in T2W imaging, DWI and ADC mapping over time?

Answer 294

Phase/time after stroke onset T2W, DWI, ADC map
Immediate post-stroke (<4-5 hrs) Isointense Hyperintense Hypointense
Acute (up to ~4 days) Hyperintense Hyperintense Hypointense
Acute-chronic (~ 4-10 days) Hyperintense Isointense Isointense
Chronic (> 10 days) Hyperintense Hypointense Hyperintense

Question 295

What is isotropic diffusion?

Answer 295

unrestricted diffusion

Question 296

What is fractional anisotropy?

Answer 296

The degree of deviation from the case of isotropic diffusion is known as fractional anisotropy (FA). The degree of FA can be calculated and assumes values between 0 and 1, where 0 represents the case of unrestricted isotropic diffusion and 1 represents the case of infinite anisotropy, for example a long thin cylinder.

Question 297

How is diffusion anisotropy used in diffusion tensor imaging?

Answer 297

diffusion tensor, which quantifies the probability of diffusion occurring with respect to distance and direction over time. The diffusion tensor is computed using combinations of the x, y and z gradients in order to calculate the ADC over a number of different directions. The direction that gives the greatest value of ADC is the principal diffusion direction.

Question 298

Where is DTI used?

Answer 298

The diffusion tensor has particular application in brain imaging. Some diseases, for example multiple sclerosis (MS), result in directional diffusion changes, such as those that occur along white matter tracts. In diffusion tensor imaging (DTI), the eigenvalues (λ1, λ2 and λ3) and eigenvectors (ε1, ε2 and ε3) of the diffusion tensor are used to create images that reflect various diffusion properties of tissue.

Diffusion tensor imaging (DTI) is particularly good for imaging neural pathways since the myelin sheath of axons provide an effective diffusion barrier. Disruption to the axonal arrangement will reduce the ordered diffusion within the tissue, and this is potentially visible as a reduction in the anisotropy.

Question 299

What are the two main methods for Perfusion MRI?

Answer 299

Dynamic contrast-enhanced MRI (DCE-MRI) and
Arterial spin labelling (ASL)
The former method is by far the most commonly used for clinical investigations.

Question 300

How is DCE-MRI performed?

Answer 300

injecting a blood-pool contrast agent (usually a Gd agent) into a vein, and then monitoring SI changes to the target tissue as the contrast agent perfuses into the tissue. The technique requires initial acquisition of baseline images prior to contrast agent injection, and then rapid acquisition of post-contrast images. Images need to be acquired with a high temporal resolution, which allows for graphical analysis of the contrast uptake to be achieved.

Question 301

What is DSC-MRI?

Answer 301

Contrast-enhanced brain perfusion studies are performed by a technique referred to as dynamic susceptibility contrast MRI (DSC-MRI). The contrast injection causes T2* (or T2) reduction in the perfused target tissues, and this results in signal loss on echo planar imaging (EPI) and T2*-weighted or T2W imaging. Quantitative parameters of interest include the:

Cerebral blood flow into the tissue (CBF)
Cerebral blood volume within the tissue (CBV)
Time taken for the blood to perfuse through the target tissue, known as the mean transit time (MTT)

Question 302

How is ASL performed?

Answer 302

In cerebral ASL, blood protons travelling towards the brain are saturated or ‘labelled’ by the application of localised radiofrequency coils (RF) excitation, using either pulsed RF or continuous RF (pulsed ASL or continuous ASL). When the saturated protons reach the brain, they perfuse into the brain tissue, which results in local SI changes. However, this effect is very small, so images are usually derived from subtraction of a labelled perfusion dataset from a non-labelled dataset.

Question 303

What is the advantage and disadvantage of ASL over DCE-MRI?

Answer 303

The advantage of ASL is that it is completely non-invasive since no contrast agent is required, but the signal-to-noise ratio is not as good as that obtained using contrast-enhanced methods.

Question 304

In the context of diffusion and perfusion MRI:
A. A diffusion sequence consists of a pair of large gradients of equal strength, either side of a 180° RF pulse
B. The higher the b-value, the less sensitive the sequence is to diffusion
C. Contrast-enhanced first pass perfusion and delayed enhancement studies in cardiac MRI usually result in infarcted left ventricular myocardium demonstrating hypointense signal at first-pass followed by hyperintense signal after a few minutes delay

Answer 304

A. True. On stationary protons, the second diffusion gradient acts to rephase the dephasing caused by the first diffusion gradient.

B. False. The larger b-value results from stronger diffusion gradients, and will render the sequence more sensitive to diffusion effects.

C. True. At first pass of the contrast bolus, an infarcted myocardial region is hypointense because the contrast agent has not been able to perfuse it. Subsequently, the same area of myocardium appears hyperintense signal (delayed enhancement) since the contrast agent has now diffused into the necrosed tissue region.

Question 305

What is the combined effect of MRI diffusion gradients on tissues with very little molecular diffusion taking place?

A. Very little effect - the second gradient counteracts the effect of the first gradient, resulting in negligible signal loss on a diffusion-weighted image Correct answer
B. A large effect - the combination of the first gradient and the second gradient causes significant signal loss - visible on a diffusion-weighted image
C. A large effect - the combination of the first gradient and the second gradient causes significant signal loss - visible on an ADC map

Answer 305

The correct answer is A.

The diffusion gradients should cancel each other out without leaving any additional dephasing due to diffusion – signal loss should therefore be minimal on a diffusion-weighted image. However, the same region would consist of hypointense signal on an ADC map.

Question 306

in fMRI is the oxygenation of venous blood higher or lower in active areas?

Answer 306

the oxygenation in the venous blood actually increases an increase in cerebral blood flow (CBF) to the region results in an increased supply that outweighs demand.

Question 307

How do the magnetic properties of Hb change when oxygenated vs deoxygenated?

Answer 307

The oxygen molecules bind to the iron atoms and change their magnetic properties. If oxygen is not present (deoxyhaemoglobin), haemoglobin is paramagnetic (weakly magnetic).

When oxygen is bound to haemoglobin (oxyhaemoglobin), the molecule effectively becomes non-magnetic (diamagnetic). This means that the magnetic properties of blood depend on the relative contributions of oxy- and deoxyhaemoglobin.

Question 308

How does the T2 of blood change when oxygenated?

Answer 308

It is known that the T2 of blood depends upon the ratio of oxyhaemoglobin to deoxyhaemoglobin. Arterial blood, which contains almost 100% oxyhaemoglobin, has a T2 of about 200 milliseconds (ms); this may be reduced to 100 ms or less in venous blood.

Question 309

What sort of sequence is used to detect T2?

Answer 309

spin-echo sequence

Question 310

what is the BOLD effect?

Answer 310

blood oxygen-level dependent (BOLD) effect.

Magnetic resonance signal contribution from blood is small compared to that from the surrounding tissues. Any change in the signal from the blood produces a relatively small overall change (less than 0.5% at 1.5 T). However, this effect can be magnified by using a sequence sensitive to T2* rather than T2. The presence of paramagnetic deoxyhaemoglobin in the venous blood vessels creates a region of magnetic field inhomogeneity in the surrounding tissues (Fig 1a). This will shorten the tissue’s T2, resulting in a reduction in the MRI signal. During activation the increase in oxyhaemoglobin will reduce the magnetic field inhomogeneity, leading to an increase in tissue T2 (Fig 1b). Since the contribution to the total MRI signal from tissue is greater, the overall effect is a 2-3% change (at 1.5 T).

Question 311

What is the main requirements for fMRI sequences?

Answer 311

The main requirement for fMRI is to have a pulse sequence that is sensitive to T2. a gradient echo (GE) with a long TR and TE gives T2 contrast.

Another requirement is to be able to scan as quickly as possible. Therefore, a fast T2*-weighted GE sequence is needed. The fastest sequence of this type is GE echo-planar imaging (EPI)

Question 312

What are the pros and cons of EPI?

Answer 312

The spatial resolution of the sequence is relatively poor at about 3 mm, compared to less than 1 mm for a standard anatomical imaging sequence. However, the trade-off is that the temporal resolution is improved. On a modern scanner, it is possible to get a stack of 36 images covering the whole brain at 3 mm resolution in about 3 seconds. Another advantage is that EPI sequences generally have a high signal-to-noise ratio (SNR); this is important when trying to look for small changes due to the BOLD effect.

Question 313

What are the difficulties in performing fMRI?

Answer 313

Performing fMRI can be very difficult, depending on the type of paradigm that will be used. Functional MRI requires scanners with high performance gradients (high gradient strengths and slew rates) in order to scan quickly.

Specialist software may also be required to run these sequences.

Moving to higher field strengths will increase the SNR and also the magnitude of the BOLD effect; these increases can be traded off to give better temporal and spatial resolution.

Question 314

What needs to be done to get usable data in fMRI?

Answer 314

In order to identify the 1-3 % change in signal statistics have to be used. This is why multiple scans have to be performed; averaging over several scans allows smaller changes to be identified as the effects of noise are reduced. The most common test is the ‘t-test’, which in simple terms, looks at the difference in the average pixel signal between two stimulus conditions relative to the spread. A large difference compared to the spread results in a high t-value

Question 315

How is the t-test data in fMRI used to create an image?

Answer 315

the pixel value is a measure of the t-value (called t-maps). A threshold is applied to these images so that only pixels with high t-values are shown.

This is then overlaid onto an anatomical image (usually the high-resolution T1 scan). The resulting colour ‘blobs’ or ‘clusters’ indicate those areas where there has been a significant change in MR signal due to the BOLD effect.

Question 316

How is fMRI used clinically?

Answer 316

Clinical applications of fMRI have not been as widespread, due to the difficulties in paradigm design, lengthy scanning time and complex analysis. However, it has been used in localising functional regions in patients with brain tumours prior to surgery. This information can be used to ensure the surgeons do not excise healthy functional tissue, which could lead to a loss of brain function. The same technique can be used for patients with temporal lobe epilepsy who may have part of their temporal lobe removed to relieve symptoms.

Question 317

Regarding fMRI:
A. Activation in an area of the brain leads to a transient decrease in blood oxygenation in that region to produce the BOLD effect
B. The BOLD response is caused by an increase in diamagnetic oxy-haemoglobin in the venous blood, reducing the magnetic field inhomogeneity around the blood vessels and increasing tissue T2*
C. The most commonly used sequence for fMRI is GE EPI
D. Event-related paradigms are very flexible, but difficult to design and analyse
E. The basic statistical test used for determining significant regions of BOLD response is called the b-test
F. Movement of the fingers of the left hand are controlled by the left-hand motor cortex

Answer 317

A. False. Although activation requires oxygen, the blood supply to the activated area increases such that the blood oxygenation rises rather than falls.

B. True. Changes in tissue T2* are the basis of the BOLD response and can be easily measured using a GE EPI sequence.

C. True. The main requirement for fMRI is to have a pulse sequence that is sensitive to T2. A GE with a long TR and TE gives T2 contrast. The fastest sequence of this type is GE EPI.

D. True. Block paradigms are the easiest to set up, with event-related paradigms being the hardest.

E. False. The t-test is used to determine the regions of significant BOLD response.

F. False. Finally, movement of the left hand fingers is controlled by the motor cortex on the contralateral side.

Question 318

What are the distinguishing features of temporal resolution in fMRI?
A. The temporal resolution of fMRI is very good
B. The temporal resolution of fMRI is worse than that of EEG
C. Functional MRI has sufficient temporal resolution to allow the identification of the onset of brain activity with high accuracy
D. The temporal resolution of fMRI is good enough to distinguish between brain activation patterns associated with different stages of stimulus processing
E. The delay between stimulus and haemodynamic response has a greater effect on fMRI temporal resolution than the fact that the response is spread over time
Submit

Answer 318

The correct answer is B.

Temporal resolution refers to the ability of a technique to distinguish events separated by short intervals of time. In this case, it can be regarded as how quickly fMRI can detect changes in brain activity in response to a stimulus; these take place over periods ranging from several milliseconds to several hundred milliseconds.

Functional MRI has poor temporal resolution of the order of 1-4 seconds. This is worse than that of other techniques; for example, EEG has a temporal resolution of 1-10 milliseconds because it measures electrical activity directly. Consequently, with fMRI it is not possible to predict the onset of brain activity with high accuracy.

The brain’s haemodynamic response imposes a fundamental limit on the accuracy of time measurement. First, the peak response is delayed by approximately 4-6 seconds with respect to the stimulus. However, this is not of great concern because the data can be adjusted to correct for the lag.

The real problem is that the response is extended over time. Temporal smoothing makes it difficult to pinpoint the precise moment of activity. Therefore, the image fMRI actually reflects an average over many seconds. This makes it very difficult to study fast neural processes because they are blurred in time.

Question 319

Why does chemical shift occur?

Answer 319

When an H atom is placed in a strong magnetic field (Fig 2), the electron orbits around the proton in such a way that it generates a small magnetic field that opposes the main field. This opposing field will shield the proton from the main magnetic field to some extent, reducing the proton’s Larmor frequency. This shift in Larmor frequency is quite small, but can be detected.

the electron shielding of the proton in an H-O bond (Fig 4) will be less than a proton in an H-C bond as the larger atom pulling the electrons closer to its nucleus, reducing the proton’s electron shielding.

This shift in Larmor frequency is known as the chemical shift.

Question 320

How is chemical shift expressed?

Answer 320

Chemical shift is often expressed as the difference from some reference frequency, expressed in parts per million (ppm). The reference frequency is that of H in tetramethylsilane (TMS)

Question 321

How are differences in FID brought out?

Answer 321

If the protons have different Larmor frequencies (i.e. different chemical shifts), then the detected FID will be the sum of the individual FIDs. The detected MR signal will contain all three frequencies. In order to extract out these frequencies, a Fourier transform is used

Question 322

What are the 2 components of an FID spectral peak?

Answer 322

The absorption spectrum
The dispersion spectrum

Question 323

What info is in the absoprtion spectrum?

Answer 323

The absorption spectrum is the most useful part.

The area under the absorption peak (also called the peak integral) is a measure of the number of protons contributing to that peak.

Chemicals present in greater concentrations will also have higher peak integrals. The full width at half-maximum (FWHM) is a measure of the spread of the peak. This is fundamentally limited by the T2 of the chemical, but will also be affected by magnetic field inhomogeneities, which contribute to T2*.

Question 324

One of the applications of fat saturation in musculoskeletal MRI is to visualise bone bruising. The fatty bone marrow is replaced by oedema.
T or F
A. If fat saturation is not correctly applied, there is a chance that healthy bone could be misinterpreted as oedema
B. The chances of seeing a fat saturation artefact does not depend on where the area being scanned is situated within the magnet
C. Artefacts due to poor fat saturation will only occur in the subcutaneous fat and not the bone marrow

Answer 324

A. T
B. F
C. F

If the fat in the bone marrow is not being saturated, it will appear bright in the image. This could be misinterpreted as bone oedema as this should not be saturated. The chances of seeing this artefact will depend on the magnetic field homogeneity, which is highest in the centre of the magnet and gets worse around the periphery.

Therefore, it is important to place the region being scanned as close to the magnet isocentre as possible. One of the ways to spot this artefact is to look at adjacent subcutaneous fat and see if that is not being saturated as well.

Question 325

what is SVS?

Answer 325

Single voxel spectroscopy (SVS) acquires an MR spectrum from a small region of tissue.

In spectroscopy, the chemical shift information is encoded in the frequency of the signal. Therefore, frequency-encoding cannot be used for spatial localisation.

In SVS, localisation is achieved using slice selection in three dimensions. This begins by defining a thick (10 to 30 mm) slice (Fig 2). Then a second slice, orthogonal to the first (Fig 3) is applied, resulting in a region that has experienced both RF pulses at the intersection (Fig 4) of the two slices.

Applying a third slice (Fig 5) orthogonal to the first two results in a cubic region at the intersection of the three slices. This is the single voxel (Fig 6) from which the MR spectrum will be obtained.

Question 326

What are 2 common sequences used in SVS?

Answer 326

Point RESolved Spectroscopy (PRESS) - The PRESS sequence is similar to a dual echo spin-echo sequence
STimulated Echo Aquisition Mode (STEAM) - The STEAM sequence uses 90° pulses to produce a stimulated echo in a similar way

The MR signal received from a PRESS sequence is usually twice that from a STEAM sequence, as the 180° pulses fully refocus the magnetisation. However, STEAM sequences can usually achieve much shorter echo times that PRESS.

Question 327

What is chemical shift imaging?

Answer 327

Chemical shift imaging (CSI) is a hybrid technique combining spectroscopy into an imaging technique. It acquires spectral data from multiple positions rather than just one.

Single voxel spectroscopy uses three intersecting slices to define the volume from which a spectrum will be obtained.

In CSI, two of these three slices are thickened , such that a flat slab is defined rather than a cube. Spatial information from this slab can be obtained by phase-encoding in two directions . This divides the slab up into an array of voxels, each with its own spectrum.

Question 328

Why does water signal need to be supressed and how is this achieved?

Answer 328

The concentration of water can be a million times greater than metabolites. The MR data is scaled to largest signal, meaning that low intensity metabolite signals may be missed. Therefore, a means of suppressing the water is required.

Common techniques for water suppression involve applying a narrow bandwidth 90° RF pulse, then a spoiler gradient to ensure that all the water magnetisation has been flipped into the transverse plane and dephased. This RF pulse gradient combination is known as CHESS (Fig 2). It is commonly used for fat saturation in standard MRI.

Question 329

How is fat suppression achieved?

Answer 329

In certain tissues, the fat signals can be as high as the water signals and therefore require some form of suppression. Since tissues containing fat and water tend to be spatially distinct, careful placement of saturation bands can be used to suppress fat signals

In situations where fat and water are not spatially distinct, special spatial-spectral RF pulses can be used. These will also suppress lipids and lactate that may be present.

Question 330

WHat are useful chemicals in neuro spectroscopy?

Answer 330

Spectroscopy has been applied in the brain for a variety of clinical conditions, including Alzheimer’s, multiple sclerosis, stroke, epilepsy and brain tumours.

a short TE spectrum will show changes in myo-inositol and glutamate/glutamine, but a longer TE spectrum will give better depiction of NAA and choline.

Question 331

Regarding chemical shift:
A. Chemical shift measured in ppm is dependent on the field strength
B. The chemical shift of H on an O-H bond is greater than H on a C-H bond

Answer 331

A. False. When expressed in units of frequency, chemical shift does depend on field strength. However, when it is expressed in ppm, it is independent of field strength.
B. True. An O nucleus (Z = 8) exerts a stronger influence on electrons than a C nucleus (Z = 6) and so the proton in an O-H bond experiences less magnetic field shielding than the proton in a C-H bond. Zero chemical shift corresponds to a situation in which the proton is shielded as much as possible. Therefore, the chemical shift for an O-H bond is greater than that for a C-H bond.

Question 332

What are the classes of MRI artefact?

Answer 332

Motion artefacts
Distortion artefacts
Radiofrequency artefacts
Data collection artefacts
Sequence-specific artefacts

Question 333

What do distortion artefacts arise from?

Answer 333

inhomogeneities in the main magnetic field (caused by the presence of metal objects or boundaries between tissues with different magnetic susceptibilities) or non-linearities in the gradient magnetic fields. They also include the displacement of fat and water due to chemical shift.

Question 334

Why does patient movement cause artefact?

Answer 334

phase- and frequency-encoding enable in-plane localisation of the magnetic resonance (MR) signal but only work if the object being scanned is stationary.

Question 335

In what plane is movement artefact most significant?

Answer 335

If the object is moving, the frequency and phase of MR signals from any point within the object will change. In frequency-encoding, a single MR signal is digitised over a very short time (a few milliseconds) and so any bulk motion is minimal; however, phase-encoding is performed over a much longer time scale where bulk motion can be significant.

Question 336

What artefact does patient motion cause?

Answer 336

Ghosting

These ‘ghosts’ are low intensity copies of the original shifted in the phase-encoding direction. The number, position and intensity of these ghosts depends on the type and severity of the movement.

Question 337

How can flow artefcts be reduced?

Answer 337

Inflow enhancement can be avoided by placing saturation bands just outside the imaging volume; magnetisation in blood flowing into the volume will be saturated, producing a low MR signal.

Question 338

How can you differentiate STIR and T2 fat saturation?

Answer 338

the STIR sequence will generally produce a more consistent fat saturation over the field-of-view (FOV). Also the contrast between tissues will be slightly different due to the inversion pulse.

Question 339

What phenomena arise from magnetic field inhomogeneity?

Answer 339

Magnetisation dephasing
Mispositioning of anatomical features

Question 340

What characteristic artefact occurs due to metal implants?

Answer 340

In SSFP sequences, magnetic field inhomogeneities produce a signal void and characteristic resonance offset artefacts, where the gradient-echoes, Hahn-echoes and stimulated echoes constructively and destructively interfere. The banding pattern demonstrates the shape and extent of homogeneity.

Question 341

Why can mispositioning occur with magnetic field inhomogeneity?

Answer 341

a localised distortion in the static field will produce a change in the total gradient, taking it away from being a straight line Magnetic resonance signal from position A will have lower Larmor frequency and will be encoded in the final image at position B superimposing on the actual signal at that position.

Question 342

How does receiver bandwidth effect distortion artefacts?

Answer 342

The extent of image distortion due to magnetic field inhomogeneity is increased by using a low receiver bandwidth.

Question 343

Why can artefacts occur due to gradient fild non-linearity?

Answer 343

constraints on the design of gradients mean that they are only linear over a certain region and then they start to ‘roll off’’ as they move away from the magnet iso-centre. In these non-linear regions, the encoded frequency will be lower than expected, resulting in an image that appears to be ‘squashed’ at the periphery of the FOV(read a full definition of this term).

Question 344

What artefact is caused by chemical shift?

Answer 344

The protons in fat are shielded by their electrons to a greater extent than those in water and therefore resonate at a lower Larmor frequency.

In the presence of a frequency-encoding gradient, fat and water that exist in the same position will appear shifted with respect to each other in the final image. This shift occurs in the frequency-encoding direction and the amount of shift (number of pixels) is greater if the frequency-encoding gradient strength is low.

Question 345

Why are there dark lines in out of phase GE images?

Answer 345

dark lines at boundaries between tissues containing fat and water (indicated by the arrow). These dark lines are absent on the in-phase image.

Where voxels contain both fat and water, the two signals can cancel each other out if a certain echo time (TE) is used.

Question 346

What is the CHESS pulse and why does it fail?

Answer 346

chemical shift selective (CHESS) pulse
narrow bandwidth that is centred over the fat peak. It flips the fat magnetisation down into the transverse plane, where it is destroyed using spoiler gradients.

should show fat as dark. However, there are regions, for example, the toes (indicated by the arrow) where the fat saturation has failed.

This is due to magnetic field inhomogeneities changing the Larmor frequency of fat protons in that region, such that the CHESS pulse is no longer centred over the fat peak.

Question 347

How do artefacts caused by spurious RF signals appear?

Answer 347

If the external RF(read a full definition of this term) signals have a specific frequency, the resulting images will possess a noise line (Fig 1) in the phase-encoding direction. The position of the line will depend upon the frequency of the spurious RF(read a full definition of this term) relative the Larmor frequency.

Question 348

What happens if there is a small FOV with tissue outside it?

Answer 348

aliasing
This occurs as the MRI signal phase shift (used to encode position in one direction for two-dimensional (2D) imaging) can only have values between -180° and +180°. A phase shift of +190° will effectively appear as a -170° phase shift, aliasing the MRI signal to the other side of the FOV. If a small FOV is required, techniques such as over-sampling can be used to eliminate this artefact.

Question 349

What is phase wrap?

Answer 349

Phase wrap can also occur in the slice direction in 3D scans. Remember that 3D imaging uses phase encoding in the slice direction. This means that the anatomy at the back of the 3D block can become aliased into slices at the front of the block.

Question 350

What artefact does a raw data point with a spurious value cause?

Answer 350

series of light and dark bands across the image
Where several spurious data points exist, the artefact in the image has a ‘herringbone’ appearance.

Spurious data points are often caused by malfunctioning equipment in the scanner room (e.g. lights, monitoring equipment, etc.) creating noise spikes that are picked up by the RF(read a full definition of this term) coil. These can also be caused by static build-up on clothing.

Question 351

What is the gibbs artefact?

Answer 351

increasing the number of data points gives a better edge. However, you will notice that there appear to be ‘waves’ superimposed on the edge. This is called edge ringing or the Gibbs artefact.

This is most evident in an image for which a reduced k-space acquisition is used (e.g. only 128 raw data points are acquired but the final image is reconstructed with 256 pixels).

Question 352

What artefacts occurs in EPI?

Answer 352

The whole of k-space is usually sampled after a single RF excitation. Sampling in the horizontal direction is very rapid, however sampling in the vertical direction is much slower. This slow sampling leads to significant distortion in areas where there are small inhomogeneities in the magnetic field, such as near the nasal cavity or around the eyes.

N/2 ghost - It is caused by misalignment of echoes due to small imbalances in the gradients.

Question 353

How can artefact be reduced in EPI?

Answer 353

parallel imaging or multi-shot EPI
Here k-space is acquired with more than one excitation or ‘shot’, but each shot samples k-space faster. While this clearly reduces the degree of distortion, it will increase the scan time.

Question 354

What artefacts can occur in FSE sequences?

Answer 354

Fast spin-echo sequences reduce the scan time by acquiring multiple lines of k-space after a single RF excitation. The number of data lines acquired per TR or echo-train length (ETL) can be very long, even to the point where the whole of k-space is obtained in a single acquisition (single-shot FSE).

However, using very high ETLs can introduce artefacts as there will be T2 relaxation over the echo train that will modulate the signal and give rise to blurring in the phase-encoding direction in the final image. This blurring will depend on the ETL and the T2 of the tissue being imaged, with the blurring being worse at high ETLs.