Regarding imaging time, which of the following increases the overall scan time?

a. Using more sequences

True

False

b. Imaging more slices

True
False

c. Defining fewer pixels

True

False

d. Reducing the TR

True

False

e. Using gadolinium

True

False

imaging time
An MRI scan consists of a number of different sequences (for example T1w, T2w, FLAIR,
Pre-contrast, post-contrast, delayed contrast). To complete one image sequence (for
example the T1weighted images), the entirety of k-space must be filled for each slice
required. The time taken to fill k-space for one slice relates to the number of phase-
encoding steps needed to cover the whole of k-space. The number of phase encoding
steps relates to the pixels in the image. For example, for a 256x256 pixel image there
must be 256 different repetitions of the sequence, each using a different phase
encoding gradient. Each different repetition takes 'TR' (time to repeat) length of time to
occur. This is usually measured in milliseconds (tens to hundreds). So the overall image
acquisition time is determined by multiplying together each of the following:

• Number of sequences - perhaps 5-10

• Number of slices - perhaps 10-50
• Number of pixels - maybe 256
• TR - around 50-1000 milliseconds

The imaging time is therefore usually around 10-60 minutes depending on exactly what
is required. Using contrast requires more sequences to be performed (pre, post, and
sometimes many different delayed sequences) so naturally will increase the scanning
time.

In the field of MRI, which of the following regarding the field of view is true?

a. A smaller FOV gives a higher resolution

True

False

b. A smaller FOV means lower gradient strengths needed

True

False

c. A smaller FOV results in higher returning signal strength

True
False

d. Fewer phase encoding steps are used for a rectangular FOV

True

False
e. A rectangular FOV has a reduced signal to noise ratio

True

False
Field of view
The physics behind changing the field of view is difficult to understand but in general,
smaller field of view:

• Higher resolution
• Higher gradient strengths needed
• The returning signal strength is lower

Rectangular fields of view can be used, for example in spine imaging. The number of
phase encoding steps is halved but the difference between each phase encoding step
is doubled such that the remaining steps cover the same overall range. There are two
problems which arise due to this technique however. Firstly, the signal to noise ratio is
reduced and secondly aliasing can occur.

Which of the following statements are true regarding MRI signal decay?

a. T1* decay occurs faster than T2* decay

True

False

b. T1 recovery occurs due to spin-lattice energy transfer

True

False

c. T2 decay occurs due to interactions with the magnetic fields of individual nuclei

True

False

d. T2* decay occurs due to magnetic inhomogeneities of the MRI scanners

magnetic field

True

False

e. T1* decay occurs due to magnetic inhomogeneities of the earth’s magnetic field

True

False
Which of the following statements are true regarding relaxation times for a 1.5T
magnet?

a. T1 is usually longer than it is in a 1 T magnet

True
False

b. T1 of fat is approximately 80ms

True

False

c. T2 of fat is approximately 260ms

True

False

d. T2 of CSF is approximately 160ms

True

False

e. T1 of CSF is approximately 2400ms

True

False

After the RF pulse...

The undergo 'relaxation' which reverts them (as a population) to their original states.
This is either spin-lattice relaxation or spin-spin relaxation.

Spin-lattice relaxation (T1 recovery, or longitudinal relaxation) - Interference from larger

nearby molecules causes energy to be redistributed. Individually, protons lose the
extra energy given to them by the RF pulse and revert to point towards the direction of
the magnetic field. As a whole, the population gradually loses M xy signal and gains
Mz signal.

Spin-spin relaxation (T2 decay, or transverse relaxation) - T2 decay occurs due to spins
dephasing from one another. This occurs due to field inhomogeneities. The
electromagnet produces a very strong magnetic field but no field is 'perfect' i.e. the
strength of the field is not completely uniform throughout. There exist miniscule
fluctuations in the strength of the field which are called 'inhomogeneities'. These
inhomogeneities affect the tiny magnetic fields produced around each spinning proton
to different extents and as the protons move around within the water of the tissues,
the different amounts of energy they have is shared out (hence spin to spin). T2 decay
times are therefore faster in tissues with very little water in them (like bone or air). The
protons in these cannot move around as much, and therefore cannot share their
energies around; they dephase faster. The higher the water content of a tissue, the
longer its T2 decay time is; hence fluid shows as high signal on T2 weighted sequences.

Image © Medical Exam Prep

T2 decay is only relevant to the Mxy plane. T1 is also relevant to the M xy plane but also
describes the signal gain in the Mz direction too.

Magnetic resonance imaging

MRI is a highly complex and unique imaging technique. To understand how it works we
have to look down to a sub-anatomic particle level.
Each atom has a nucleus and each nucleus contains a specific number of protons
(plus/minus neutrons) depending on which element it is. Nuclei of elements which have
an odd number of protons in their nuclei (1,3,5,7 etc.) possess a property called nuclear
magnetic resonance. Hydrogen nuclei which consist only of one single proton are
present within water in almost everything we are looking to image with MRI and so for
the most part, it is the behaviour hydrogen nuclei with which we are most interested in
seeking to understand how MRI works. Of all the atoms in a human body around 63%
(by number) are hydrogen.

Image © Medical Exam Prep

Each proton has a property called 'spin'; they spin like the earth spins on its north-
south axis. The axis is also called the dipole. The direction of the dipole is represented
by an arrow. Normally within a substance there is no overall direction but when that
substance is placed inside a magnetic field the dipoles all align along the field-lines of
the magnetic field pointing either in the same direction or the opposite direction. If the
dipole points along the direction of the field it is described either as 'spin-up' or
parallel. If it points against it, it is termed 'spin-down' or antiparallel.

Image © Medical Exam Prep

The electromagnet of an MRI machine is very powerful. Due to the high currents
passing through the coil the apparatus would overheat if it was not supercooled.
Cooling is achieved by bathing the coil in liquid helium at temperatures of -270oC or
4Kelvin. Normally helium is a gas. At 5Kelvin it forms a liquid.
Magnetic fields are measured in either 'Gauss' or 'Tesla' where 1 Gauss is the strength
of the earth's own magnetic field, 1Tesla is equivalent to 10,000Gauss and a typical
diagnostic MRI scanner is 1.5 or 3Tesla in strength.
How many dipoles are in the spin-up state?
The images we produce in MRI depend on the dipoles that are in the spin-down state.
The majority of spins align with along the direction of the magnetic field but a few align
in the opposite direction. At body temperature, about 5 hydrogens out of every one
million align parallel but against the direction of the magnetic field. This is a tiny
number and yet it is the behaviour of these protons which eventually form an MRI
image.

Which of the following statements are true regarding vectors?

a. A vector has both magnitude and direction

True

False

b. Different vectors cannot be added to one another

True

False

c. The Mz vector describes the degree of magnetisation pointing parallel to the

bore of the magnet
True

False

d. The Mxy vector points towards the ceiling

True
False

e. Protons do not spin when fully in the Mz plane

True

False

Vectors
A vector is a force in a particular direction. Different vectors with different magnitudes
and directions or angles can be added (or subtracted) to give an overall vector. In MRI
physics we deal with two vectors primarily; the Mz vector, and the Mxy vector.
Regarding MRI safety, which of the following is true?

a. The gradient magnetic fields are always on

True

False

b. The gradient magnetic fields pose a projectile risk

True
False

c. Aluminium equipment is not a projectile risk

True
 False

d. Patients with a cardiac pacemaker cannot have MRI scans

True
False

e. The gradient magnetic field can cause peripheral nerve stimulation

True

False

MRI hazards and safety

Although there is no ionising radiation to contend with during an MRI scan, there are
still plenty of hazards to be aware of.

• The magnetic field. The magnetic field is always on. This is because it is
easier/better for the machine to remain on permanently than to repeatedly have
to supercool the coil to the extent needed in order for it to function correctly.
The immediate concerns regarding the static magnetic field relate to risk from
ferromagnetic objects entering the room becoming projectiles. Equipment used
within the MRI room must be made from MR-safe, non-ferrous materials.
Austenitic stainless steel (a type of stainless steel) is MR-safe, as is aluminium.
Cardiac pacemakers are of significant concern due to the risk, not just of lead
movement, but also of abnormal current induction in the leads leading to
arrhythmias. Some types of pacemaker (or neural stimulators) are MR
conditional (an update concept of MR-safe, with limits on the strength of the
magnetic field they have been tested to), although should still be switched off
temporarily for the duration of the scan.

• The gradient magnetic fields. These are made to apply in various different
directions for varying amounts of time during the scan. Notable side effects
include peripheral nerve stimulation giving the sensation of tingling, but also a
metallic taste in the mouth, vertigo and nausea. Involuntary muscular
contractions and cardiac arrythmias have also been reported. These effects are
all more prevalent when the background magnetic field strength is more
significant (for example in a 3T magnet as opposed to a 1.5T one). Obviously the
shorter the duration of the variations occurring, the lower the risk of side effects
occurring. The risk of the gradient magnetic fields to the developing foetus are
theoretical however it is to guard against the effects of this aspect of the scan
that scans are usually ideally not performed on or by patients/staff in their first
trimester of pregnancy (where declared), though most follow-up studies do not
 show noticeable effects for the foetus.

The second aspect of the gradient magnetic fields which causes a problem is the
noise incurred when the gradients are applied. Having an MRI scan is a very
noisy experience. The noise comes from switching on and off the gradient
magnetic fields. Patients are given ear defenders to guard against hearing
damage, and for their own comfort.

Which of the following reduce scan time for an MRI scan?

a. Short TR

True
False

b. Increased phase matrix

True

False

c. Increased number of excitations

True

False

d. Reduced SNR

True
False

e. Use of gadolinium

True

False

Scan time
Scanning time is an important parameter. Longer scans are associated with
significantly more patient discomfort which in turn leads to motion artefact and in
many cases, early termination of the scan without the necessary sequences having
been acquired. The following all reduce scan time but many of these must be traded
off to preserve image quality:
 • Short TR – this causes a reduction in SNR and causes the images to be more
heavily T1 weighted
• Reduced number of phase encoding steps – i.e. reduced phase matrix. This
reduces the resolution
• Reduced number of excitations. This reduces the SNR and increases the effect
from motion artefact if there is any.

Which of the following will improve image contrast in MRI imaging?

a. Improving SNR

True
False

b. Gadolinium

True

False

c. SPIOs

True

False

d. Reducing the FOV

True
False

e. Using a larger matrix

True

False

Which of the following improve spatial resolution in MRI imaging?

a. Using smaller pixels

 True
False

b. Using a smaller FOV

True

False
c. Using a larger matrix

True

False

d. Using thinner slices

True

False

e. Using gadolinium

True

False

Contrast
Contrast between different tissues within an MRI image depends on the comparison of
the SNRs within those tissues. Contrast does therefore depend, to an extent, on the
SNR. Other techniques can help improve the contrast between structures as well:

• Fat suppression – This is a useful technique for some specific purposes. Fat
suppression is performed with post-gadolinium images such that the high signal
on the T1 images used to show the contrast medium, can be assured not to
have come from fat. STIR images are fat supressed.
• Contrast media – such as gadolinium or SPIOs help promote image contrast by
returning more signal from structures which contain the contrast.
• Magnetisation transfer constant – this involves including in the sequence used,
some RF pulses delivered at specific off-resonant frequencies. These are
targeted at protons bound within big molecules (instead of in water molecules),
 and they help reduce the signal from the big molecules so that the signal from
the water molecules is better appreciated

Spatial resolution
The spatial resolution of an MRI image relates to the size of the pixels being used. This
itself depends on the FOV (field of view) which is selected on the scout images, and on
the matrix selected on the computer. Smaller pixels allow smaller structures to be
distinguished hence smaller pixels improve spatial resolution. The following all improve
spatial resolution:

• Smaller pixels
• Smaller FOV
• Larger matrix
• Use of a local coil (e.g. a head coil, as opposed to the large coil contained within
the gantry)
• Thinner slices (reduce the partial volume effect).

Which of the following statements are true regarding k-space?

a. K-space is a table of spatial frequencies

True
False

b. Each horizontal line represents data from each frequency encoded gradient
applied

True

False

c. Each horizontal line is converted to a line in the image by Fourier transform

True
False

d. In k-space the brightness of each dot represents the amplitude of each

frequency

True
False
e. In a normal spin-echo sequence, many lines of k-space can be acquired during
each TR

True

False

Only one line of k-space can be acquired in a TR during a normal spin-echo sequence.
More can be acquired during a turbo spin echo sequence though.

Spatial encoding
This is the process of encoding the returning signal such that we can work out where in
the body the signal has come from. This has to be done in all three directions (x, y and
z).

K-space**
K-space is the theoretical graph which has to be filled in for each slice before the image
can be produced. It is filled in line-by-line by successive repetitions of each sequence.
For example, in a spin-echo sequence, the first TR might be used to collect the bottom
line along k-space, then the phase encoding gradient is changed for the next TR such
that the next line up can be read and added. Only once the entirety of the k-space
graph has been collected for that slice can a Fourier transform be applied to the whole
thing, the result of which is the diagnostic image for that slice.
 Image © Medical Exam Prep

It is important to note that each area in k-space does not correspond to individual
pixels in the final image. Each point in k-space contains frequency and phase encoded
information which relate to the picture as a whole.

The centre of k-space contains spatial frequencies relating to basic image contrast.

The periphery of k-space contains spatial frequencies relating to the detail within the
image, the edges or boundaries between two tissue types.
Image © Medical Exam Prep
**The letter 'k' itself comes from the universal term given to the concept of a
'wavenumber'. A wavenumber is 1 divided by the wavelength i.e. it is the number of
wavelengths per unit distance.

Which of the following regarding the radiofrequency pulses used in MRI scanning is
true?

a. A 90o pulse lasts for one and a half minutes

True
False

b. A 180o pulse turns net magnetisation to point in the opposite direction

True

False

c. Both 90o and 180o pulses cause spin rephasing

True

False

d. After a 180o pulse there are more spins pointing antiparallel than parallel to M z
True

False

e. Both 90o and 180o pulses are emitted at the same frequency; the Larmor
frequency
True

False

Radiofrequency (RF) pulses

As well as the large coil of the electromagnet, there are also separate coils either
contained within the apparatus of the scanner, or placed directly on the patient (head
coils, body coils). The coils are designed such that they produce a magnetic field which
oscillates to and fro at a frequency of around 63.9MHz (in a 1.5T machine). Firstly, this
is in the radiofrequency part of the electromagnetic spectrum, and secondly this
frequency is chosen because it is the Larmor frequency for water. When a pulse of RF
energy is applied, the effect on tissue exposed to this radiofrequency energy is twofold:

• Some of the spin-up protons gain enough energy to 'flip' to become spin-down
protons. This causes a change in the Mz vector. The Mz vector can increase,
 decrease or disappear completely depending on the strength and duration of
the RF pulse.
• The protons in the tissue become are forced to precess in phase with one
another. This means that their hitherto random distribution (like clocks in
different time zones) in the Mxy plane becomes a single magnetic vector in that
plane, spinning around like clocks in synch. Their synchronisation is responsible
for a net magnetic vector being produced in the Mxy plane.

Image © Medical Exam Prep

90o pulse
A 90o pulse is one which tips the net magnetisation (by 90o (ie into the xy plane), and
also causes re-phasing such that the spins are all precessing in phase with one another.

180o pulse
A 180o pulse is one which tips the net magnetisation by 180o (ie into the opposite
direction to the Mz direction), and also causes re-phasing such that the spins are all
precessing in phase with one another. Prior to a 180o pulse there is a tiny excess in
protons pointing parallel to the direction of the magnetic field compared to pointing
anti-parallel. After the 180othis excess is reversed; there are slightly more protons
pointing antiparallel.
Which of the following statements are true regarding phase encoding?

a. Phase encoding typically occurs in the y-axis

True

False

b. Phase encoding gradients are applied during signal readout

True
False

c. The phase shift is proportional to the position from front to back

True

False

d. Phase encoding must be repeated as many times as there are pixels in that
direction

True

False

e. Within k-space, each horizontal line represents signal from a different phase
encoding gradient

True

False

patial encoding
This is the process of encoding the returning signal such that we can work out where in
the body the signal has come from. This has to be done in all three directions (x, y and
z).

Phase encoding - this process encodes one of the other two axes, typically the y-axis
(front to back). After the slice has been selected, a gradient is applied in the y-direction
which alters the phases of the spins from the front of the patient to the back. This is
akin to time-zones around the world. Time is still passing with the same speed but the
clock hands are pointing to different numbers. The gradient applied can be a strong
gradient (where the phase shifts a lot from front to back), or a shallow gradient (where
the phase shifts only a small amount from front to back), or anything in between. The
sequence is repeated with many different gradient strengths, one after another and
each time, when the signal is read out, comparing the phase of the waves between
successive gradient strengths allows us (a computer) to deduce where along the y-axis
those spins gave signal. The phase shift is proportional to the position of the spins
generating the signal, along the y-axis. The resultant data is added line by line to
something called k-space which is a theoretical graph, and the whole of k-space must
be filled in before an image can be produces from it. To collect enough data there must
be as many repetitions, all with different strength phase gradients applied as there are
number of pixels from front to back.
Each line added to k-space represents the signal collected under different strengths of
phase-encoding gradient. There might be a strong gradient where the magnitude of
the phase shift changes significantly over the width of the slice, or a mild gradient
where the magnitude of the phase shift changes only slightly over the width of the
slice. The line added to k-space does not itself relate to position along the y-axis, only
that the phase shifts undergone by the different frequencies within that signal are
proportional to position along the y-axis. Phase encoding is complicated!
 Image © Medical Exam Prep

k-space**
k-space is the theoretical graph which has to be filled in for each slice before the image
can be produced. It is filled in line-by-line by successive repetitions of each sequence.
For example, in a spin-echo sequence, the first TR might be used to collect the bottom
line along k-space, then the phase encoding gradient is changed for the next TR such
that the next line up can be read and added. Only once the entirety of the k-space
graph has been collected for that slice can a Fourier transform be applied to the whole
thing, the result of which is the diagnostic image for that slice.

It is important to note that each area in k-space does not correspond to individual
pixels in the final image. Each point in k-space contains frequency and phase encoded
information which relate to the picture as a whole.
**The letter 'k' itself comes from the universal term given to the concept of a
'wavenumber'. A wavenumber is 1 divided by the wavelength i.e. it is the number of
wavelengths per unit distance.

Regarding artefacts in MRI imaging, which of the following is true?

a. The resonance of hydrogen nuclei is the same throughout the body

True

False

b. Metallic joint replacements can cause susceptibility artefacts

True

False

c. Motion artefact occurs in the frequency encoding direction

True

False

d. Moving blood returns little or no signal on spin-echo sequences

True

False

e. Gadolinium has super-paramagnetic properties

True

False

Motion artefact
Frequency encoding takes place very quickly, within a fraction of a millisecond and as
such is not as severely affected by motion as phase encoding, which takes place over a
matter of hundreds-thousands of milliseconds. This means that on the resultant image,
the effects of patient motion are only really noticeable in the phase-encoding direction.
For structures which pulsate (e.g. the aorta), there can be multiple images of the same
structure all lined up on top of one another along the phase encoding gradient. This is
known as ghosting.

Magnetic susceptibility
The metal of a joint replacement distorts the magnetic field around the joint and alters
the signal returning from that area. The affected area is usually black with anatomical
distortion at the point where the signal becomes recognisable.

Chemical shift artefact

The resonant frequency of hydrogen atoms is different for different hydrogen atoms
depending on the substance they find themselves in. This is known as the Larmor
frequency. The Larmor frequency of fat is slightly higher than that of water. The
composition of the sequences and pulses makes the assumption that the Larmor
frequencies are identical throughout the body. When this ceases to be the case it is
seen as a black or white band either side of a structure. For example, where the kidney
meets the perinephric fat, a black line on one side, and a white line on the other side
may be found. This is because the adjacent fat, resonating at a slightly higher
frequency is mis-interpreted by the frequency-encoding gradient as being located
slightly to the side of where it truly is. This effect is known as the chemical shift artefact.

Contrast agents - exogenous

The composition of gadolinium ions includes seven unpaired electrons. These unpaired
electrons make them strongly paramagnetic. Unfortunately, gadolinium is very toxic so
must be attached or chelated to other compounds which bind the gadolinium tightly,
allowing it first to dissolve in water and then to be excreted by the body shortly
following administration.

Its paramagnetic properties cause significant spin dephasing in any spins in its vicinity.
It reduces both T1 and T2 however it has a far greater effect on T1 than it does on T2,
hence all IV contrast sequences are performed as T1 weighted sequences.

Gd-DTPA does not cross the normal blood-brain barrier so can be used to discern
when this has been pathologically breached.

Contrast – endogenous
Arterial spin labelling – this is a way of providing vascular contrast without using an
exogenous substance. It relies on the different T1 re-growth times of blood as
compared to surrounding tissue. Two images are acquired, one where the signal from
blood is nulled by spin inversion, and a second control image. When the two images
are combined, an image of perfusion can be obtained.

Time of flight imaging – This technique relies on the blood in the slice downstream
from the slice of interest is targeted and the spins excited by RF pulses. As the blood in
this upstream slice flows into the target slice downstream, the spins from the blood
show bright against the slice which hasn’t received these initial RF pulses.

Regarding MRS, which of the following is true?

a. Stands for Magnetic Resonance Susceptibility

True
False

b. It is a qualitative test

True
False

c. The peaks correspond to shifts of the resonant frequencies

True

False
d. NAA is usually pathological if low

True

False

e. Choline is usually pathological if high

True
False

Magnetic resonance spectroscopy (MRS)

This is a quantitative technique. Within a given voxel, the frequencies returned can be
examined and deconstructed to show the content of the voxel. The ‘peaks’ measured
relate to the degree to which the resonant frequency has shifted which in turn depends
on the position and number of hydrogens present within the molecules in the voxel.
The peaks correspond to different known compounds such as N-acetylaspartate,
choline and creatine. The ratios of these peaks to one another is used as an indicator
of pathology. In almost all conditions which use MRS, a low NAA and a high choline
level suggest pathology.
Regarding aliasing, the following reduce the likelihood of aliasing occurring?

a. Selecting a larger field of view

True

False

b. Applying the phase encoding gradient in the shortest diameter needed

True
False

c. Using pre-saturation bands

True

False

d. Using gadolinium

True

False

e. Leaving a short gap between sequences

True

False

Field of view
The physics behind changing the field of view is difficult to understand but in general,
smaller field of view:

• Higher resolution
• Higher gradient strengths needed
• The returning signal strength is lower

Rectangular fields of view can be used, for example in spine imaging. The number of
phase encoding steps is halved but the difference between each phase encoding step
is doubled such that the remaining steps cover the same overall range. There are two
problems which arise due to this technique however. Firstly, the signal to noise ratio is
reduced and secondly aliasing can occur.
Aliasing, or wrap-around artefact
This occurs when the body part being imaged is bigger than the field of view selected.
The signal coming from the part of the body outside the field of view is misinterpreted
as coming from the other side of the image. Although this can theoretically occur in
both the phase encoding and the frequency encoding gradients, in practice the phase
encoding gradient is only applied to the field of view. Aliasing occurs in the frequency
encoding gradient if the returning signal is not sampled frequently enough. To avoid
this artefact the signal must be sampled at at least twice the wavelength of the highest
frequency. This (sampling twice per wavelength) is known as the Nyquist criteria. If the
signal is under-sampled then pixels from outside of the field of view are accidentally
mapped onto areas within the field of view. The classic example is the nose of a patient
appearing as an occipital mass.

Aliasing can be avoided by:

• Using a larger field of view

• Changing over the phase and frequency encoding gradients (if only a problem in
one direction)
• Applying a pre-saturation band to the area outside the field of view which
removes signal except from within the field of view
• Using software which corrects for the artefact.

Which of the following reduce spatial resolution in MRI imaging?

a. Using smaller pixels

True

False

b. Using a larger FOV

True

False

c. Using a smaller matrix

True
 False

d. Using thinner slices

True
False

e. Using SPIOs

True

False

Regarding contrast agents used in MRI scanning, which of the following is true?

a. Gadolinium has diamagnetic properties

True
False

b. Gadolinium is harmless in its chelated form

True

False

c. Gadolinium reduces T2 signal

True

False

d. Gd-DTPA crosses the blood-brain barrier easily

True

False

e. Gd-DTPA is water soluble

True
False
Contrast agents
The composition of gadolinium (Gd) ions includes seven unpaired electrons. These
unpaired electrons make them strongly paramagnetic. Unfortunately, gadolinium is
very toxic so must be attached or chelated to other compounds which bind the
gadolinium tightly, allowing it first to dissolve in water and then to be excreted by the
body shortly following administration.

Its paramagnetic properties cause significant spin dephasing in any spins in its vicinity.
It reduces both T1 and T2 however it has a far greater effect on T1 than it does on T2,
hence all IV contrast sequences are performed as T1 weighted sequences. The
reduction of the spin-lattice relaxation time (T1) caused by the gadolinium means that
tissues containing the contrast return a higher signal on T1 weighted sequences.

Gd-DTPA does not cross the normal blood-brain barrier so can be used to discern
when this has been pathologically breached.

SPIOs (superparamagnetic iron oxide) can also be used as an exogenous contrast

medium. Iron is ferromagnetic but the miniscule nanoparticles of iron oxide are too
small to be truly ferromagnetic. Instead they also exhibit paramagnetic effects.

Direction of Relative
Property Typical material
polarisation strength

Diamagnetic Opposite -10 Water, Biological tissues

Gadolinium, Manganese,
Paramagnetic Same +1
Copper

Superparamagnetism Same +5000 SPIOs, Haemosiderin

Ferromagnetic Same >10,000 Iron, Steel

• Diamagnetic materials have paired electrons and have a negative (albeit weak)
susceptibility to surrounding magnetic fields. They align against the direction of
polarisation but revert to a random distribution if the magnetic field is removed.
• Ferromagnetic materials have at least one unpaired electron. They are strongly
attracted to align in the same direction as the magnetic field and, once the
magnetic field is removed, remain that way aligned afterwards, retaining also
magnetisation of their own.
• Paramagnetic materials are weakly attracted to align in the same direction as
the magnetic field. They also have at least one unpaired electron.
Which of the following statements are true regarding k-space?

a. The centre of k-space relates to rough image contrast

True

False

b. The periphery of k-space relates to signal from gadolinium

True

False

c. K-space is not needed for CT scanning, only MRI

True

False

d. Fourier transforms are used in both CT scanning and MRI

True
False

e. K-space must be filled in for both the phase encoding gradient and the
frequency encoding gradient, then the two are averaged

True

False

K-space is filled in once per slice with data from each repetition of the sequence
occurring under each new phase encoding gradient.

Spatial encoding
This is the process of encoding the returning signal such that we can work out where in
the body the signal has come from. This has to be done in all three directions (x, y and
z).

K-space**
K-space is the theoretical graph which has to be filled in for each slice before the image
can be produced. It is filled in line-by-line by successive repetitions of each sequence.
For example, in a spin-echo sequence, the first TR might be used to collect the bottom
line along k-space, then the phase encoding gradient is changed for the next TR such
that the next line up can be read and added. Only once the entirety of the k-space
graph has been collected for that slice can a Fourier transform be applied to the whole
thing, the result of which is the diagnostic image for that slice.

Image © Medical Exam Prep

It is important to note that each area in k-space does not correspond to individual
pixels in the final image. Each point in k-space contains frequency and phase encoded
information which relate to the picture as a whole.

The centre of k-space contains spatial frequencies relating to basic image contrast.

The periphery of k-space contains spatial frequencies relating to the detail within the
image, the edges or boundaries between two tissue types.
Image © Medical Exam Prep
**The letter 'k' itself comes from the universal term given to the concept of a
'wavenumber'. A wavenumber is 1 divided by the wavelength i.e. it is the number of
wavelengths per unit distance.

Which of the following statements are true regarding MRI?

a. It is only possible to produce images using hydrogen nuclei

True

False

b. Faradays law of induction describes that in a closed circuit, changes in the

magnetic flux can induce a current

True

False

c. The gyromagnetic ratio of a hydrogen atom is 42.6kHzT -1

True

False

d. Spin-lattice relaxation is the decay of transverse magnetism

True

False

e. The time constant of T2 is the same as for T1 in pure water

True
False

Which of the following statements are true regarding T1 weighted sequences?

a. Fat returns high signal

True
 False

b. Water returns high signal

True
False

c. A typical time to echo is around 15ns

True

False

d. A typical time to repeat is 500ms

True

False

e. The TE and TR are shorter than for T2 weighted sequences

True

False

Sequence TR (ms) TE (ms)

T1w (both short) 500 15
T2w (both long) 4000 90
FLAIR (both very long) 9000 115
Proton density (long TR,
2000 15
short TE)

• TR is always longer than TE

• A short TR is less than 500ms
• A long TR is more than 1500ms, around 3x the length of a short TR
• A short TE is less than 30ms
• A short TE is more than 90ms, around 3x the length of a short TE.

Sequences
To produce a signal from the tissue of interest, different strength pulses are used for
different durations, in different orders. Each set of pulses is known as a 'sequence'. The
sequences are designed for different purposes for example:
 • T1 weighted sequence - fat, blood, melanin, proteinaceous fluid and gadolinium
are all high in signal on a T1 weighted sequence. This sequence can be used with
or without fat suppression. Once the signal from fat has been supressed,
gadolinium based IV contrast can also be given and imaged using this sequence.
• T2 weighted sequence - fluid and fat are both high in signal
• PD sequence - proton density sequence, similar to T2, used a lot for MSK
imaging and neuroradiology.
• Diffusion weighted sequences - These are concerned with the movement of
water molecules within tissues. Anisotropy (An-eye-sot-rop-y) is a property which
describes how non-linear the diffusion of water through a tissue is. In highly
organised/structured tissues such as cardiac muscle fibres, or neuronal
networks, there is a much stronger anisotropic diffusion of water through the
tissue along the lines of the muscle fibres or neurones. Certain pathologies (in
brain imaging: infarction, demyelination, pus, and some cancers) can disrupt this
anisotropy and 'restrict diffusion'. DW sequences must always be coupled with
an 'ADC' map (Apparent Diffusion Coefficient). An ADC map is produced by
eliminating the effect of T2 from the DW sequence; DW sequences are
inherently very heavily T2 weighted. If something is high in signal intensity on
both sequences, this is due to its T2 effects (T2-shine through), and not its
diffusion properties.
• Flow-related sequences - there are a few different varieties of these. Some use
injected contrast, and some rely on the movement of blood around the field of
view.

Which of the following regarding the 90o radiofrequency pulses used in MRI scanning is
true?

a. The RF pulses are set to emit the frequency of the Larmor frequency of water
hydrogen nuclei

True
False

b. There are three main effects of the RF pulse on the protons

True

False
c. Protons rephase during the duration of the RF pulse
 True
False

d. Magnetisation in the Mz vector may decrease during an RF pulse

True

False

e. The RF pulse produces a net magnetic vector in the Mxy plane

True

False

Radiofrequency (RF) pulses

As well as the large coil of the electromagnet, there are also separate coils either
contained within the apparatus of the scanner, or placed directly on the patient (head
coils, body coils). The coils are designed such that they produce a magnetic field which
oscillates to and fro at a frequency of around 63.87MHz (in a 1.5T machine). Firstly, this
is in the radiofrequency part of the electromagnetic spectrum, and secondly this
frequency is chosen because it is the Larmor frequency for water. When a pulse of RF
energy is applied, the effect on tissue exposed to this radiofrequency energy is twofold:

• Some of the spin-up protons gain enough energy to 'flip' to become spin-down
protons. This causes a change in the Mz vector. The Mz vector can increase,
decrease or disappear completely depending on the strength and duration of
the RF pulse.
• The protons in the tissue become are forced to precess in phase with one
another. This means that their hitherto random distribution (like clocks in
different time zones) in the Mxy plane becomes a single magnetic vector in that
plane, spinning around like clocks in synch. Their synchronisation is responsible
for a net magnetic vector being produced in the Mxy plane.
Which of the following statements are true or false regarding MRI sequences?

a. All sequences must be performed for each complete scan

True

False

b. Sequence choice depends on the clinical indication

True
False

c. Contrast sequences are usually T1 weighted

True

False

d. Fat is high in signal on both T1 and T2 weighted sequences

True
False

e. Proton density sequences are useful for imaging joints

True
 False

Sequences
To produce a signal from the tissue of interest, different strength pulses are used for
different durations, in different orders. Each set of pulses is known as a 'sequence'. The
sequences are designed for different purposes for example:

• T1 weighted sequence - fat, blood, melanin, proteinaceous fluid and gadolinium

are all high in signal on a T1 weighted sequence. This sequence can be used with
or without fat suppression. Once the signal from fat has been supressed,
gadolinium based IV contrast can also be given and imaged using this sequence.
• T2 weighted sequence - fluid and fat are both high in signal
• PD sequence - proton density sequence, similar to T2, used a lot for MSK
imaging
• Diffusion weighted sequences - These are concerned with the movement of
water molecules within tissues. Anisotropy (An-eye-sot-rop-y) is a property which
describes how non-linear the diffusion of water through a tissue is. In highly
organised/structured tissues such as cardiac muscle fibres, or neuronal
networks, there is a much stronger anisotropic diffusion of water through the
tissue along the lines of the muscle fibres or neurones. Certain pathologies (in
brain imaging: infarction, demyelination, pus, and some cancers) can disrupt this
anisotropy and 'restrict diffusion'. DW sequences must always be coupled with
an 'ADC' map (Apparent Diffusion Coefficient). An ADC map is produced by
eliminating the effect of T2 from the DW sequence; DW sequences are
inherently very heavily T2 weighted. If something is high in signal intensity on
both sequences, this is due to its T2 effects (T2-shine through), and not its
diffusion properties.
• Flow-related sequences - there are a few different varieties of these. Some use
injected contrast, and some rely on the movement of blood around the field of
view.

To produce these sequences, different combinations of radiofrequency pulses are

used. Spin-echo pulse sequence: This is the workhorse of MRI sequences.
 Image © Medical Exam Prep

Gradient-echo sequence: Useful if the sequence needs to have a short TR. This
sequence suffers from magnetic inhomogeneities because there is no 180 o pulse to
'refocus' the spins which have suffered from this effect. Consequently, any T2 effect
seen is referred to as T2* (T2 star). Local susceptibility artefact can be used to
diagnostic advantage in many circumstances though. In cerebral imaging, if there are
microhaemorrhages, the presence of excess iron causes spin dephasing leading to
'blooming' artefact.

Spin-echo Gradient-echo

180o refocussing pulse No 180 o pulse

First flip angle used is 90o Any angle can be used

Relatively slow sequence Much faster sequence

 T2 weighting is 'true' T2* weighting occurs, not T2 weighting

Long TR can be used Works for shorter TRs

Why does a 180 deg pulse rephase the spins?

Imagine two people, one fast and one slow, running a race around a circular track. At
some point a bystander blows a whistle and the two people turn around again and race
back towards the start. At some point the two people become level again. The analogy
is imperfect because the spins do not reverse their direction of spinning. In fact, the
analogy is closer to a race being run on a disk which is then inverted by 180o. The spins
carry on spinning but now they are chasing one another back again; rephasing.

In the field of MRI, regarding the concept of spin relaxation, which of the following is
true?

a. A stronger magnet causes longer T1 values

True

False

b. A stronger magnet causes shorter T2 times

True

False

c. T2 is usually longer than T1

True
False

d. T1 and T2 are usually a similar length to one another

True

False

e. T2 is only relevant to decay in the Mxy plane

True
False
Radiofrequency (RF) pulses
As well as the large coil of the electromagnet, there are also separate coils either
contained within the apparatus of the scanner, or placed directly on the patient (head
coils, body coils). The coils are designed such that they produce a magnetic field which
oscillates to and fro at a frequency of around 63.9 MHz (in a 1.5T machine). Firstly, this
is in the radiofrequency part of the electromagnetic spectrum, and secondly this
frequency is chosen because it is the Larmor frequency for water. When a pulse of RF
energy is applied, the effect on tissue exposed to this radiofrequency energy is twofold:

• Some of the spin-up protons gain enough energy to 'flip' to become spin-down
protons. This causes a change in the Mz vector. The Mz vector can increase,
decrease or disappear completely depending on the strength and duration of
the RF pulse.
• The protons in the tissue become are forced to precess in phase with one
another. This means that their hitherto random distribution (like clocks in
different time zones) in the Mxy plane becomes a single magnetic vector in that
plane, spinning around like clocks in synch. Their synchronisation is responsible
for a net magnetic vector being produced in the Mxy plane.

They undergo 'relaxation' which reverts them (as a population) to their original states.
This is either spin-lattice relaxation or spin-spin relaxation.

Spin-lattice relaxation (T1 recovery, or longitudinal relaxation) - Interference from larger

nearby molecules causes energy to be redistributed. Individually, protons lose the
extra energy given to them by the RF pulse and revert to point towards the direction of
the magnetic field. As a whole, the population gradually loses M xy signal and gains
Mz signal.

Spin-spin relaxation (T2 decay, or transverse relaxation) - T2 decay occurs due to spins
dephasing from one another. This occurs due to field inhomogeneities. The
electromagnet produces a very strong magnetic field but no field is 'perfect' i.e. the
strength of the field is not completely uniform throughout. There exist miniscule
fluctuations in the strength of the field which are called 'inhomogeneities'. These
inhomogeneities affect the tiny magnetic fields produced around each spinning proton
to different extents and as the protons move around within the water of the tissues,
the different amounts of energy they have is shared out (hence spin to spin). T2 decay
times are therefore faster in tissues with very little water in them (like bone or air). The
protons in these cannot move around as much, and therefore cannot share their
energies around; they dephase faster. The higher the water content of a tissue, the
longer its T2 decay time is; hence fluid shows as high signal on T2 weighted sequences.

Image © Medical Exam Prep

T2 decay is only relevant to the Mxy plane. T1 is also relevant to the M xy plane but also
describes the signal gain in the Mz direction too.

Tissue Approximate T1 (ms) Approximate T2 (ms)

Bone Long Short

CSF 2000 150

Grey matter 800 100

White matter 650 90

Spleen 400 60

Kidney 550 60
 Liver 400 40

Fat 250 80

• Stronger magnets cause higher T1 values

• The magnet strength does not change T2
• T2 always has to be shorter than T1

Which of the following statements are true regarding proton density weighted
sequences?

a. Water is poorly seen on this sequence

True

False

b. Usually has a TE around 15ms

True

False

c. This is one of the only sequences where bone returns high signal

True

False

d. Has a long TR

True

False
e. Is used a lot for neuroradiology

True

False
 Sequence TR (ms) TE (ms)
T1w (both short) 500 15
T2w (both long) 4000 90
FLAIR (both very long) 9000 115
Proton density (long TR,
2000 15
short TE)

• TR is always longer than TE

• A short TR is less than 500ms
• A long TR is more than 1500ms, around 3x the length of a short TR
• A short TE is less than 30ms
• A long TE is more than 90ms, around 3x the length of a short TE.

Sequences
To produce a signal from the tissue of interest, different strength pulses are used for
different durations, in different orders. Each set of pulses is known as a 'sequence'. The
sequences are designed for different purposes for example:

• T1 weighted sequence - fat, blood, melanin, proteinaceous fluid and gadolinium

are all high in signal on a T1 weighted sequence. This sequence can be used with
or without fat suppression. Once the signal from fat has been supressed,
gadolinium based IV contrast can also be given and imaged using this sequence.
• T2 weighted sequence - fluid and fat are both high in signal
• PD sequence - proton density sequence, similar to T2, used a lot for MSK
imaging and neuroradiology.
• Diffusion weighted sequences - These are concerned with the movement of
water molecules within tissues. Anisotropy (An-eye-sot-rop-y) is a property which
describes how non-linear the diffusion of water through a tissue is. In highly
organised/structured tissues such as cardiac muscle fibres, or neuronal
networks, there is a much stronger anisotropic diffusion of water through the
tissue along the lines of the muscle fibres or neurones. Certain pathologies (in
brain imaging: infarction, demyelination, pus, and some cancers) can disrupt this
anisotropy and 'restrict diffusion'. DW sequences must always be coupled with
an 'ADC' map (Apparent Diffusion Coefficient). An ADC map is produced by
eliminating the effect of T2 from the DW sequence; DW sequences are
inherently very heavily T2 weighted. If something is high in signal intensity on
both sequences, this is due to its T2 effects (T2-shine through), and not its
diffusion properties.
 • Flow-related sequences - there are a few different varieties of these. Some use
injected contrast, and some rely on the movement of blood around the field of
view.

Regarding MRI safety considerations during pregnancy, which of the following is true?

a. Gadolinium should be avoided during the first trimester

True

False

b. A foetus undergoing an MRI scan at 20 weeks is at significant risk of hearing

damage

True

False

c. Chelated gadolinium is teratogenic

True

False

d. Chelated gadolinium is strongly excreted in breast milk

True

False

e. 24 hours following administration, gadolinium is undetectable in breast milk

True
False

MRI hazards and safety

Although there is no ionising radiation to contend with during an MRI scan, there are
still plenty of hazards to be aware of.
Gradient magnetic fields
The risk of the gradient magnetic fields to the developing foetus are theoretical
however it is to guard against the effects of this aspect of the scan that scans are
usually ideally not performed on or by patients/staff in their first trimester of
pregnancy (where declared), though most follow-up studies do not show noticeable
effects for the foetus. Foetal perception of noise begins around week 18. Although MRI
machines are very noisy, studies have not shown there to be an association with
hearing problems in children whose mother’s underwent MRI scans during gestation,
indeed foetal MRI as a prognostic tool is becoming more widespread, but
understandably there is usually still a reasonable degree of concern regarding infant
hearing in scans performed around this stage. In general, if it is feasible to wait until
after delivery for an MRI scan to be performed then this is usually deemed the best
course of action.

Gadolinium in pregnancy
Gadolinium also crosses the placenta and is excreted by the foetus into the amniotic
fluid. The chelated form of gadolinium in itself poses no risk to the foetus however the
free form of gadolinium is teratogenic and the longer the foetus is exposed to the
chelated gadolinium administered for the scan, the greater the risk of each molecule of
gadolinium becoming dissociated from its chelate. For this reason therefore,
gadolinium should only be administered in a patient known to be pregnant, if there is
likely to be significant diagnostic benefit from the resultant images.

Considerations during lactation

Less than 1% of injected iodine or gadolinium is excreted into breast milk and less than
1% of this is then absorbed by the baby's intestinal tract. Although these amounts are
considered acceptably low as to not necessitate any breast-feeding cessation advice
many families are understandably over-cautious. It is accepted that contrast agents are
undetectable in blood 24 hours after administration and so any mother wishing to
'pump-and-dump' despite the negligible risks need not do so for longer than a day.
Care should, however, be given to counsel the family that it is possible for inadvertent
weaning to occur even after short intervals and this should be factored in to any
decision made by the family.

In the field of MRI, regarding the concept of spin relaxation, which of the following is
true?
a. Spin-lattice interactions cause longitudinal relaxation

True

False

b. Spin-spin interactions are caused by interference from larger nearby molecules

True
False

c. T2* decay is caused by magnetic field inhomogeneities

True

False

d. Both T1 recovery and T2 decay are concerned with regrowth of the M z vector
True

False

e. Transverse relaxation has a greater effect in more rigid tissues

True

False

adiofrequency (RF) pulses

As well as the large coil of the electromagnet, there are also separate coils either
contained within the apparatus of the scanner, or placed directly on the patient (head
coils, body coils). The coils are designed such that they produce a magnetic field which
oscillates to and fro at a frequency of around 63.87MHz (in a 1.5T machine). Firstly, this
is in the radiofrequency part of the electromagnetic spectrum, and secondly this
frequency is chosen because it is the Larmor frequency for water. When a pulse of RF
energy is applied, the effect on tissue exposed to this radiofrequency energy is twofold:

• Some of the spin-up protons gain enough energy to 'flip' to become spin-down
protons. This causes a change in the Mz vector. The Mz vector can increase,
decrease or disappear completely depending on the strength and duration of
the RF pulse.
• The protons in the tissue become are forced to precess in phase with one
another. This means that their hitherto random distribution (like clocks in
different time zones) in the Mxy plane becomes a single magnetic vector in that
 plane, spinning around like clocks in synch. Their synchronisation is responsible
for a net magnetic vector being produced in the Mxy plane.

They undergo 'relaxation' which reverts them (as a population) to their original states.
This is either spin-lattice relaxation or spin-spin relaxation.

Spin-lattice relaxation (T1 recovery, or longitudinal relaxation) - Interference from larger

nearby molecules causes energy to be redistributed. Individually, protons lose the
extra energy given to them by the RF pulse and revert to point towards the direction of
the magnetic field. As a whole, the population gradually loses M xy signal and gains
Mz signal.

Spin-spin relaxation (T2 decay, and T2* decay, or transverse relaxation) - T2 decay
occurs due to spins dephasing from one another causing a reduction in the height of
the Mz vector. T2* is when this occurs due to the effects of field inhomogeneities, in
addition to the effects of normal molecule interactions. The electromagnet produces a
very strong magnetic field but no field is 'perfect' i.e. the strength of the field is not
completely uniform throughout. There exist miniscule fluctuations in the strength of
the field which are called 'inhomogeneities'. These inhomogeneities affect the tiny
magnetic fields produced around each spinning proton to different extents and as the
protons move around within the water of the tissues, the different amounts of energy
they have is shared out (hence spin to spin). T2 decay times are therefore faster in
tissues with very little water in them (like bone or air). The protons in these cannot
move around as much, and therefore cannot share their energies around; they
dephase faster. The higher the water content of a tissue, the longer its T2 decay time is;
hence fluid shows as high signal on T2 weighted sequences.
 Image © Medical Exam Prep

T2 decay is only relevant to the Mxy plane. T1 is also relevant to the M xy plane but also
describes the signal gain in the Mz direction too.

Tissue Approximate T1 (ms) Approximate T2 (ms)

Bone Long Short

CSF 2000 150

Grey matter 800 100

White matter 650 90

Spleen 400 60

Kidney 550 60

Liver 400 40

Fat 250 80

hich of the following statements regarding MRI are correct?

a. MRI depends on a property called nuclear magnetic resonance

True

False

b. A helium nucleus consists of only one proton

True

False

c. Applying a magnetic field to protons causes them to spin

True

False

d. A dipole is a tiny magnetic field around a spinning proton

True

False

e. When the dipoles align, the protons stop spinning

True

False

In the field of MRI, regarding quenching the electromagnet, which of the following is
correct?

a. The electromagnet is too strong to leave on when not in use

True

False

b. Quenching should only be performed in emergency situations

True

False

c. It is not possible to re-use the supercooled helium following a quench

 True
False

d. The valves in the door to an MRI room are to equalise the air pressure in the
case of a quench

True

False

e. Spare replacement helium should be stored on site in case a quench is

necessary

True
False

Magnetic resonance imaging

MRI is a highly complex and unique imaging technique. To understand how it works we
have to look down to a sub-anatomic particle level.

Each atom has a nucleus and each nucleus contains a specific number of protons
(plus/minus neutrons) depending on which element it is. Nuclei of elements which have
an odd number of protons in their nuclei (1,3,5,7 etc.) possess a property called nuclear
magnetic resonance. Hydrogen nuclei which consist only of one single proton are
present within water in almost everything we are looking to image with MRI and so for
the most part, it is the behaviour hydrogen nuclei with which we are most interested in
seeking to understand how MRI works. Of all the atoms in a human body around 63%
(by number) are hydrogen.
 Image © Medical Exam Prep

Each proton has a property called 'spin'; they spin like the earth spins on its north-
south axis. The axis is also called the dipole. The direction of the dipole is represented
by an arrow. Normally within a substance there is no overall direction but when that
substance is placed inside a magnetic field the dipoles all align along the field-lines of
the magnetic field pointing either in the same direction or the opposite direction. If the
dipole points along the direction of the field it is described either as 'spin-up' or
parallel. If it points against it, it is termed 'spin-down' or antiparallel.
 Image © Medical Exam Prep

The electromagnet of an MRI machine is very powerful. Due to the high currents
passing through the coil the apparatus would overheat if it was not supercooled.
Cooling is achieved by bathing the coil in liquid helium at temperatures of -270oC or
4Kelvin. Normally helium is a gas. At 5Kelvin it forms a liquid.
Magnetic fields are measured in either 'Gauss' or 'Tesla' where 1 Gauss is the strength
of the earth's own magnetic field, 1Tesla is equivalent to 10,000Gauss and a typical
diagnostic MRI scanner is 1.5 or 3Tesla in strength.

How many dipoles are in the spin-up state?

The images we produce in MRI depend on the dipoles that are in the spin-down state.
The majority of spins align with along the direction of the magnetic field but a few align
in the opposite direction. At body temperature, about 5 hydrogens out of every one
million align parallel but against the direction of the magnetic field. This is a tiny
number and yet it is the behaviour of these protons which eventually form an MRI
image.
It takes slightly more energy for a proton to exist in the spin-down state meaning that
there are very slightly fewer protons in the spin-down state than the spin-up. The
images we produce in MRI depend on this difference. At body temperature, the
difference is only about 5 hydrogens out of every one million align. This is a tiny
number and yet it is the behaviour of these protons which eventually form an MRI
image.

Precession
'Precession' is the term used to describe the spinning action of a proton which is, still
spinning on its axis, but moving so that the ends of the axis seem to trace a circle.
Individually, these spinning precessing protons generate minute magnetic fields
themselves. Under the influence of a strong external magnetic field, these minute
magnetic fields line up in parallel with the field (either with it or against it). The result is
that the 'net magnetisation' is zero. This is not to say that in the context of a static
magnetic field the protons stop precessing; they don't. Only that they precess in such a
way that their tiny magnetic fields, when summed together create a net magnetisation
of zero. To create signal, radiofrequency (RF) pulses are applied to the patient which
cause the axes to tip-over. When the RF pulses are switched off, the precession forms a
spiral path to return their original positions.
 Image © Medical Exam Prep

A rotating frame of reference

This concept is difficult to visualise. A similar real-life example using a fairground
merry-go-round would be the difference between watching the horses go up and down
from the side compared to watching them go up and down while standing on the ride
yourself. In the first scenario you have to account for the fact that the horses are
moving past you as well as up and down but in the second it is only their up and down
motion that you perceive.
 Image © Medical Exam Prep

A rotating frame of reference is useful in understanding MRI as it allows us to draw in

two dimensions what is occurring in three dimensions. When an RF pulse is applied,
and then removed, the spins 'tip-over' and then, while still precessing, rise back up
again. Considering the vectors at play, the z-component of the spin becomes bigger,
and the xy-component reduces as the spin tilts back upwards again. In a rotating frame
of reference, this decay of the xy-component is seen as a sinusoidal pulse wave.

Terminology
Bo - This is the direction of the magnetic field.
Mz - When an RF pulse is removed, the z-component of the spins increases or 'grows'.
Phrases such as 're-growth of the Mz vector' are used.
Mxy - When an RF pulse is removed, the spins lose strength in the M xy plane. This is
described as 'decay of the Mxy plane'.
The Larmor frequency
The frequency (in MHz) with which protons precess is dependent on a constant called
the gyromagnetic ratio, which is different for the protons in different nuclei.

The Larmor frequency (ω) = gyromagnetic ratio (γ) x magnetic field strengh (B o)

Echo planar imaging

This is a very fast version of gradient echo sequence. During the application of the
frequency-encoding gradient, the direction of the gradient is oscillated between
positive and negative very very fast. Each time it switches a phase encoding gradient is
also applied very briefly and a gradient echo is produced. This fills k-space very fast
meaning the whole head can be imaged in a matter of only a couple of sounds.
Unfortunately, the resolution and SNR are all very poor using this method.

Diffusion weighted imaging.

DWI uses echo planar imaging to form an appreciation of the way in which water
molecules move about within the slice. Static spins within a slice have their signals
cancelled out by successive gradient pulses but spins which move while the gradients
are being applied suffer a phase shift. If the spins are moving isotropicly (i.e. in all
directions due to Brownian motion of molecules, without constraint of tissue structure)
then the phase shifts again cancel due to interference from one another but if there is
a structure to the tissue such as that created by neuronal networks within the brain
then the water diffuses preferentially (or anisotropically) along the direction of the
neurones and this produces a measurable difference in signal. The same signal must
be acquired in all three planes in order to build up a diffusion weighted image. These
are then combined to give an overall picture of the amount to which diffusion is
restricted in each voxel. The diffusion of water within damaged, oedematous tissues,
becomes restricted from its usual linear fashion and shows up as high signal.

Historically DWI and EPI imaging was achieved on gradient echo sequences but these
days different manufacturers use different sequences, either gradient or spin echo
sequences to produce diffusion weighting.
Noise in MRI
Noise in MRI comes from a variety of different sources.

• Movement of molecules within the body

• Electrical noise within the coils and electromagnet itself
• RF interference from electrical equipment within the room (usually this would be
removed where possible once its presence is noted)

SNR in MRI
There are a number of different parameters which can improve SNR in MRI:

• Stronger magnetic field (e.g. using a 3T machine instead of a 1.5T one). In a

stronger magnetic field, more of the spins are aligned along the axis of the
magnetic field so there is a stronger signal to be tipped, dephased, rephased etc.
• Phased coil arrays – as opposed to more simple coils. Phased array coils allow
multiple different signals to be received at once, each on its own channel, for
example using a 16-channel coil is better than using a 2-channel coil.
• Short TE - At shorter TEs there is less spin dephasing so the magnetisation in the
Mxy plane is greater, the signal is stronger and the SNR is better. But, at very
short TEs, the T2 signal is reduced because it has not had sufficient time to
develop.
• Long TR – A longer TRs allows the Mz vector to regrow as much as possible,
producing a stronger signal. This effect is however, limited by the deleterious
effect on the T1 signal. The longer the TR is allowed to go on, the greater the
spin-lattice spin dephasing.
• Using a larger flip angle – Larger flip angles increase the magnetisation
generated in the Mxy plane and allow for a shorter TE to be used. This is also
helpful in reducing scan times as well as improving SNR.
• Reducing the receiver bandwidth – the receiver bandwidth is the expanse of
frequencies collected during signal acquisition. Although this means the scan
time has to be longer in order to allow enough time to collect all the frequencies
across a greater number of smaller bandwidths, the SNR does improve.
Which of the following are true regarding the selection of slices during an MRI scan?

a. Spatial encoding is done during installation of the MRI machine

True

False

b. Slice selection is usually performed in the z-axis

True
False

c. Slice selection is achieved by causing the magnetic field to vary from one end of
the scanner to the other

True

False

d. Slice select relies on varying the Larmor frequency from one end of the patient
to another

True
False

e. The Larmor frequency is always at least 42.6MHz or higher (for a 1.5T magnet)

True

False

Spatial encoding
This is the process of encoding the returning signal such that we can work out where in
the body the signal has come from. This has to be done in all three directions (x, y and
z).

Slice select - this process encodes which slice we are imaging, i.e. the z-axis. The
electromagnet strength is varied linearly from one end of the patient to the other. For
example, if we are imaging the brain then at the vertex the magnetic field is made to be
very slightly stronger and at the base of the skull it is made to be very slightly weaker.
Or the variation can be across the whole body. Variation across shorter lengths allows
for thinner slices to be selected. The strength of the magnetic field determines the
precession frequency of the spins within that slice so to 'select the slice' to image, the
RF pulses are applied at the Larmor frequency according to the slice needed. The
magnetic field can be made to vary through the use of additional coils which act
against the background of the existing magnetic field.

Which of the following are true regarding frequency encoding during an MRI scan?

a. Frequency encoding typically occurs in the y-axis

True
False

b. The strength of the magnetic field is made to vary from left to right
 True
False

c. The frequency encoding gradient is applied at the same time as slice selection

True

False
d. A Fourier transform is used to show the different frequencies as amplitudes

True

False

e. Frequency encoding must be repeated to fill in k-space line by line

True

False
To produce these sequences, different combinations of radiofrequency pulses are
used.

Spin-echo pulse sequence: This is the workhorse of MRI sequences.

 Image © Medical Exam Prep

Gradient-echo sequence: Useful if the sequence needs to have a short TR. This
sequence suffers from magnetic inhomogeneities because there is no 180 o pulse to
'refocus' the spins which have suffered from this effect. Consequently, any T2 effect
seen is referred to as T2* (T2 star). Local susceptibility artefact can be used to
diagnostic advantage in many circumstances though. In cerebral imaging, if there are
microhaemorrhages, the presence of excess iron causes spin dephasing leading to
'blooming' artefact.

Spin-echo Gradient-echo

180o refocussing pulse No 180o pulse

First flip angle used is 90o Any angle can be used

Relatively slow sequence Much faster sequence

T2 weighting is 'true' T2* weighting occurs, not T2 weighting

Long TR can be used Works for shorter TRs

Sequence TR (ms) TE (ms)

T1w (both short) 500 15
T2w (both long) 4000 90
FLAIR (both very long) 9000 115
Proton density (long TR,
2000 15
short TE)

• TR is always longer than TE

• A short TR is less than 500ms
• A long TR is more than 1500ms, around 3x the length of a short TR
• A short TE is less than 30ms
• A long TE is more than 90ms, around 3x the length of a short TE.

Sequences
To produce a signal from the tissue of interest, different strength pulses are used for
different durations, in different orders. Each set of pulses is known as a 'sequence'. The
sequences are designed for different purposes for example:

• T1 weighted sequence - fat, blood, melanin, proteinaceous fluid and gadolinium

are all high in signal on a T1 weighted sequence. This sequence can be used with
or without fat suppression. Once the signal from fat has been supressed,
gadolinium based IV contrast can also be given and imaged using this sequence.
• T2 weighted sequence - fluid and fat are both high in signal
• PD sequence - proton density sequence, similar to T2, used a lot for MSK
imaging and neuroradiology.
• Diffusion weighted sequences - These are concerned with the movement of
water molecules within tissues. Anisotropy (An-eye-sot-rop-y) is a property which
describes how non-linear the diffusion of water through a tissue is. In highly
organised/structured tissues such as cardiac muscle fibres, or neuronal
networks, there is a much stronger anisotropic diffusion of water through the
tissue along the lines of the muscle fibres or neurones. Certain pathologies (in
brain imaging: infarction, demyelination, pus, and some cancers) can disrupt this
 anisotropy and 'restrict diffusion'. DW sequences must always be coupled with
an 'ADC' map (Apparent Diffusion Coefficient). An ADC map is produced by
eliminating the effect of T2 from the DW sequence; DW sequences are
inherently very heavily T2 weighted. If something is high in signal intensity on
both sequences, this is due to its T2 effects (T2-shine through), and not its
diffusion properties.
• Flow-related sequences - there are a few different varieties of these. Some use
injected contrast, and some rely on the movement of blood around the field of
view.

Inversion recovery
This sequence is a little bit like a spin-echo sequence but with an extra gradient bolted
onto the front. To begin with, a 180o inversion sequence is used which flips the
Mz vector upside down. Different tissues revert at slightly different rates, with the
Mz vector gradually growing. Within each tissue type, there becomes a point in time
(specific to that tissue type) where the Mz vector is zero.
 Image © Medical Exam Prep

By choosing the tissue type (e.g. water for FLAIR, or fat for STIR imaging), the normal
spin-echo sequence can be made to begin at exactly the point at which the Mz signal
from that tissue type is ‘nulled’. This means that that tissue type then does not
participate in the remainder of the sequence; signal comes only from other tissues.
This is referred to as that tissue type being ‘nulled’.
 Image © Medical Exam Prep

FLAIR stands for Fluid Attenuation Inversion Recovery. FLAIR images show fluid as
black.

STIR stands for Short T1 Inversion Recovery. (Fat has a short T1). STIR images cause the
signal from fat to be supressed.

ADC stands for Apparent Diffusion Coefficient.

• FLAIR stands for Fluid Attenuation Inversion Recovery. FLAIR images show fluid
as black.
• STIR stands for Short T1 Inversion Recovery. (Fat has a short T1). STIR images
cause the signal from fat to be supressed.
• PD sequence - proton density sequence, similar to T2, used a lot for MSK
imaging and neuroradiology.
 • Diffusion weighted sequences - These are concerned with the movement of
water molecules within tissues. Anisotropy (An-eye-sot-rop-y) is a property which
describes how non-linear the diffusion of water through a tissue is. In highly
organised/structured tissues such as cardiac muscle fibres, or neuronal
networks, there is a much stronger anisotropic diffusion of water through the
tissue along the lines of the muscle fibres or neurones. Certain pathologies (in
brain imaging: infarction, demyelination, pus, and some cancers) can disrupt this
anisotropy and 'restrict diffusion'. DW sequences must always be coupled with
an 'ADC' map (Apparent Diffusion Coefficient). An ADC map is produced by
eliminating the effect of T2 from the DW sequence; DW sequences are
inherently very heavily T2 weighted. If something is high in signal intensity on
both sequences, this is due to its T2 effects (T2-shine through), and not its
diffusion properties.

Contrast agents
The composition of gadolinium (Gd) ions includes seven unpaired electrons. These
unpaired electrons make them strongly paramagnetic. Unfortunately, gadolinium is
very toxic so must be attached or chelated to other compounds which bind the
gadolinium tightly, allowing it first to dissolve in water and then to be excreted by the
body shortly following administration.

Its paramagnetic properties cause significant spin dephasing in any spins in its vicinity.
It reduces both T1 and T2 however it has a far greater effect on T1 than it does on T2,
hence all IV contrast sequences are performed as T1 weighted sequences.

Gd-DTPA does not cross the normal blood-brain barrier so can be used to discern
when this has been pathologically breached.

SPIOs (superparamagnetic iron oxide) can also be used as an exogenous contrast

medium. Iron is ferromagnetic but the miniscule nanoparticles of iron oxide are too
small to be truly ferromagnetic. Instead they also exhibit paramagnetic effects.

Direction of Relative
Property Typical material
polarisation strength

Diamagnetic Opposite -10 Water, Biological tissues

 Gadolinium, Manganese,
Paramagnetic Same +1
Copper

Superparamagnetism Same +5000 SPIOs, Haemosiderin

Ferromagnetic Same >10,000 Iron, Steel

• Diamagnetic materials have paired electrons and have a negative (albeit weak)
susceptibility to surrounding magnetic fields. They align against the direction of
polarisation but revert to a random distribution if the magnetic field is removed.
• Ferromagnetic materials have at least one unpaired electron. They are strongly
attracted to align in the same direction as the magnetic field and, once the
magnetic field is removed, remain that way aligned afterwards, retaining also
magnetisation of their own.
• Paramagnetic materials are weakly attracted to align in the same direction as
the magnetic field. They also have at least one unpaired electron.

In the field of MRI, regarding the concept of spin relaxation, which of the following is
true?

a. Spin-lattice interactions cause longitudinal relaxation

True

False

b. Spin-spin interactions are caused by interference from larger nearby molecules

True

False
c. T2* decay is caused by magnetic field inhomogeneities

True

False

d. Both T1 recovery and T2 decay are concerned with regrowth of the M z vector
True

False
e. Transverse relaxation has a greater effect in more rigid tissues
 True
False

Radiofrequency (RF) pulses

As well as the large coil of the electromagnet, there are also separate coils either
contained within the apparatus of the scanner, or placed directly on the patient (head
coils, body coils). The coils are designed such that they produce a magnetic field which
oscillates to and fro at a frequency of around 63.87MHz (in a 1.5T machine). Firstly, this
is in the radiofrequency part of the electromagnetic spectrum, and secondly this
frequency is chosen because it is the Larmor frequency for water. When a pulse of RF
energy is applied, the effect on tissue exposed to this radiofrequency energy is twofold:

• Some of the spin-up protons gain enough energy to 'flip' to become spin-down
protons. This causes a change in the Mz vector. The Mz vector can increase,
decrease or disappear completely depending on the strength and duration of
the RF pulse.
• The protons in the tissue become are forced to precess in phase with one
another. This means that their hitherto random distribution (like clocks in
different time zones) in the Mxy plane becomes a single magnetic vector in that
plane, spinning around like clocks in synch. Their synchronisation is responsible
for a net magnetic vector being produced in the Mxy plane.

They undergo 'relaxation' which reverts them (as a population) to their original states.
This is either spin-lattice relaxation or spin-spin relaxation.

Spin-lattice relaxation (T1 recovery, or longitudinal relaxation) - Interference from larger

nearby molecules causes energy to be redistributed. Individually, protons lose the
extra energy given to them by the RF pulse and revert to point towards the direction of
the magnetic field. As a whole, the population gradually loses M xy signal and gains
Mz signal.

Spin-spin relaxation (T2 decay, and T2* decay, or transverse relaxation) - T2 decay
occurs due to spins dephasing from one another causing a reduction in the height of
the Mz vector. T2* is when this occurs due to the effects of field inhomogeneities, in
addition to the effects of normal molecule interactions. The electromagnet produces a
very strong magnetic field but no field is 'perfect' i.e. the strength of the field is not
completely uniform throughout. There exist miniscule fluctuations in the strength of
the field which are called 'inhomogeneities'. These inhomogeneities affect the tiny
magnetic fields produced around each spinning proton to different extents and as the
protons move around within the water of the tissues, the different amounts of energy
they have is shared out (hence spin to spin). T2 decay times are therefore faster in
tissues with very little water in them (like bone or air). The protons in these cannot
move around as much, and therefore cannot share their energies around; they
dephase faster. The higher the water content of a tissue, the longer its T2 decay time is;
hence fluid shows as high signal on T2 weighted sequences.

Image © Medical Exam Prep

T2 decay is only relevant to the Mxy plane. T1 is also relevant to the M xy plane but also
describes the signal gain in the Mz direction too.

Tissue Approximate T1 (ms) Approximate T2 (ms)

Bone Long Short

CSF 2000 150

Grey matter 800 100

White matter 650 90

Spleen 400 60

Kidney 550 60

Liver 400 40

Fat 250 80

Which of the following statements are true regarding vectors?

a. A vector has both magnitude and direction

True

False

b. Different vectors cannot be added to one another

True

False

c. The Mz vector describes the degree of magnetisation pointing parallel to the

bore of the magnet
True

False

d. The Mxy vector points towards the ceiling

True
False

e. Protons do not spin when fully in the Mz plane

True

False
Vectors
A vector is a force in a particular direction. Different vectors with different magnitudes
and directions or angles can be added (or subtracted) to give an overall vector. In MRI
physics we deal with two vectors primarily; the Mz vector, and the Mxy vector.

Regarding MRI safety, which of the following is true?

a. The gradient magnetic fields are always on

True

False

b. The gradient magnetic fields pose a projectile risk

 True
False

c. Aluminium equipment is not a projectile risk

True

False
d. Patients with a cardiac pacemaker cannot have MRI scans

True

False

e. The gradient magnetic field can cause peripheral nerve stimulation

True

False

MRI hazards and safety

Although there is no ionising radiation to contend with during an MRI scan, there are
still plenty of hazards to be aware of.

• The magnetic field. The magnetic field is always on. This is because it is
easier/better for the machine to remain on permanently than to repeatedly have
to supercool the coil to the extent needed in order for it to function correctly.
The immediate concerns regarding the static magnetic field relate to risk from
ferromagnetic objects entering the room becoming projectiles. Equipment used
within the MRI room must be made from MR-safe, non-ferrous materials.
Austenitic stainless steel (a type of stainless steel) is MR-safe, as is aluminium.
Cardiac pacemakers are of significant concern due to the risk, not just of lead
movement, but also of abnormal current induction in the leads leading to
arrhythmias. Some types of pacemaker (or neural stimulators) are MR
conditional (an update concept of MR-safe, with limits on the strength of the
magnetic field they have been tested to), although should still be switched off
temporarily for the duration of the scan.

• The gradient magnetic fields. These are made to apply in various different
directions for varying amounts of time during the scan. Notable side effects
include peripheral nerve stimulation giving the sensation of tingling, but also a
metallic taste in the mouth, vertigo and nausea. Involuntary muscular
 contractions and cardiac arrythmias have also been reported. These effects are
all more prevalent when the background magnetic field strength is more
significant (for example in a 3T magnet as opposed to a 1.5T one). Obviously the
shorter the duration of the variations occurring, the lower the risk of side effects
occurring. The risk of the gradient magnetic fields to the developing foetus are
theoretical however it is to guard against the effects of this aspect of the scan
that scans are usually ideally not performed on or by patients/staff in their first
trimester of pregnancy (where declared), though most follow-up studies do not
show noticeable effects for the foetus.

The second aspect of the gradient magnetic fields which causes a problem is the
noise incurred when the gradients are applied. Having an MRI scan is a very
noisy experience. The noise comes from switching on and off the gradient
magnetic fields. Patients are given ear defenders to guard against hearing
damage, and for their own comfort.

Regarding artefacts in MRI imaging, which of the following is true?

a. The resonance of hydrogen nuclei is the same throughout the body

True

False

b. Metallic joint replacements can cause susceptibility artefacts

True

False

c. Motion artefact occurs in the frequency encoding direction

True

False

d. Moving blood returns little or no signal on spin-echo sequences

True

False
e. Gadolinium has super-paramagnetic properties

True

False

Motion artefact
Frequency encoding takes place very quickly, within a fraction of a millisecond and as
such is not as severely affected by motion as phase encoding, which takes place over a
matter of hundreds-thousands of milliseconds. This means that on the resultant image,
the effects of patient motion are only really noticeable in the phase-encoding direction.
For structures which pulsate (e.g. the aorta), there can be multiple images of the same
structure all lined up on top of one another along the phase encoding gradient. This is
known as ghosting.

Magnetic susceptibility
The metal of a joint replacement distorts the magnetic field around the joint and alters
the signal returning from that area. The affected area is usually black with anatomical
distortion at the point where the signal becomes recognisable.

Chemical shift artefact

The resonant frequency of hydrogen atoms is different for different hydrogen atoms
depending on the substance they find themselves in. This is known as the Larmor
frequency. The Larmor frequency of fat is slightly higher than that of water. The
composition of the sequences and pulses makes the assumption that the Larmor
frequencies are identical throughout the body. When this ceases to be the case it is
seen as a black or white band either side of a structure. For example, where the kidney
meets the perinephric fat, a black line on one side, and a white line on the other side
may be found. This is because the adjacent fat, resonating at a slightly higher
frequency is mis-interpreted by the frequency-encoding gradient as being located
slightly to the side of where it truly is. This effect is known as the chemical shift artefact.

Contrast agents - exogenous

The composition of gadolinium ions includes seven unpaired electrons. These unpaired
electrons make them strongly paramagnetic. Unfortunately, gadolinium is very toxic so
must be attached or chelated to other compounds which bind the gadolinium tightly,
allowing it first to dissolve in water and then to be excreted by the body shortly
following administration.

Its paramagnetic properties cause significant spin dephasing in any spins in its vicinity.
It reduces both T1 and T2 however it has a far greater effect on T1 than it does on T2,
hence all IV contrast sequences are performed as T1 weighted sequences.

Gd-DTPA does not cross the normal blood-brain barrier so can be used to discern
when this has been pathologically breached.

Contrast – endogenous
Arterial spin labelling – this is a way of providing vascular contrast without using an
exogenous substance. It relies on the different T1 re-growth times of blood as
compared to surrounding tissue. Two images are acquired, one where the signal from
blood is nulled by spin inversion, and a second control image. When the two images
are combined, an image of perfusion can be obtained.

Time of flight imaging – This technique relies on the blood in the slice downstream
from the slice of interest is targeted and the spins excited by RF pulses. As the blood in
this upstream slice flows into the target slice downstream, the spins from the blood
show bright against the slice which hasn’t received these initial RF pulses.

Regarding MRI safety, which of the following is true?

a. Gradient magnetic fields can cause heating

True

False

b. Gradient magnetic fields can cause arrhythmias in patients who do not have
pacemakers

True
False

c. The use of multiple RF pulses during sequences creates a lot of acoustic noise

True

False
d. The patient’s weight is needed to calculate the SAR

True

False

e. Burns can occur at the point of contacts between patients and wires

True
False

MRI hazards and safety

Although there is no ionising radiation to contend with during an MRI scan, there are
still plenty of hazards to be aware of.

• The magnetic field. The magnetic field is always on. This is because it is
easier/better for the machine to remain on permanently than to repeatedly have
to supercool the coil to the extent needed in order for it to function correctly.
The immediate concerns regarding the static magnetic field relate to risk from
ferromagnetic objects entering the room becoming projectiles. Equipment used
within the MRI room must be made from MR-safe, non-ferrous materials.
Austenitic stainless steel (a type of stainless steel) is MR-safe, as is aluminium.
Cardiac pacemakers are of significant concern due to the risk, not just of lead
movement, but also of abnormal current induction in the leads leading to
arrhythmias. Some types of pacemaker (or neural stimulators) are MR
conditional (an update concept of MR-safe, with limits on the strength of the
magnetic field they have been tested to), although should still be switched off
temporarily for the duration of the scan.

• The gradient magnetic fields. These are made to apply in various different
directions for varying amounts of time during the scan. Notable side effects
include peripheral nerve stimulation giving the sensation of tingling, but also a
metallic taste in the mouth, vertigo and nausea. Involuntary muscular
contractions and cardiac arrythmias have also been reported. These effects are
all more prevalent when the background magnetic field strength is more
significant (for example in a 3T magnet as opposed to a 1.5T one). Obviously the
shorter the duration of the variations occurring, the lower the risk of side effects
occurring. The risk of the gradient magnetic fields to the developing foetus are
theoretical however it is to guard against the effects of this aspect of the scan
that scans are usually ideally not performed on or by patients/staff in their first
trimester of pregnancy (where declared), though most follow-up studies do not
show noticeable effects for the foetus.
 The second aspect of the gradient magnetic fields which causes a problem is the
noise incurred when the gradients are applied. Having an MRI scan is a very
noisy experience. The noise comes from switching on and off the gradient
magnetic fields. Patients are given ear defenders to guard against hearing
damage, and for their own comfort.

• RF pulses. The risk from the RF pulses is of heating. In a whole-body sense, there
are limits applied to scans which curb the SAR (specific absorption ratio) allowed
to be given to a patient during the scan. The ratio limits depend on the weight of
the patient, hence patients are weighed prior to scanning in order to calculate
the limit allowed. Furthermore, there is a risk of contact-point heating. This can
occur if there is metal in contact with the patient (non-ferrous), and for this
reason patients are usually advised to change into a patient gown prior to the
scan. Contact-point heating can also occur in ‘closed-loop’ scenarios. Where a
closed-loop exists, a small current can run within it during the application of an
RF pulse. This includes, for example, if the patient has crossed their arms; the
loop of the arms and body would count as a loop. At the point of skin contact
there is inefficiency in the conduction of this current which manifests as heat
and can lead to burns (even second/third degree burns in rare cases). For this
reason, care is taken to ensure, not just that the patient has not crossed their
arms or legs during the scan, but also that where monitoring leads are
necessary, they do not cross over the patient but rather run straight down the
bore of the scanner. It is also the reason why it is not possible for caregivers (e.g.
a parent) to hold the hand of their child during paediatric scans, or other care-
related scenarios.

• The fringe magnetic field. The strength of the magnetic field diminishes
exponentially away from the bore. In the fringes of the magnetic field there is far
less of a projectile risk however there is still a risk to electrical items such as
watches, phones and items containing chips such as cards, especially contactless
cards. For this reason, public access to the area around the magnet where the
magnetic field strength is still above 0.5mT is restricted even if it is not within the
confines of a designated ‘inner controlled area’. Hence there will be areas
outside the inner controlled area where public access is restricted; because the
field strength is still greater than 0.5mT. The inner controlled area must contain
all parts of the field in excess of 3mT.
Regarding aliasing, the following reduce the likelihood of aliasing occurring?

a. Selecting a larger field of view

True

False

b. Applying the phase encoding gradient in the shortest diameter needed

True

False

c. Using pre-saturation bands

True

False
d. Using gadolinium

True

False

e. Leaving a short gap between sequences

True
False

Field of view
The physics behind changing the field of view is difficult to understand but in general,
smaller field of view:

• Higher resolution
• Higher gradient strengths needed
• The returning signal strength is lower

Rectangular fields of view can be used, for example in spine imaging. The number of
phase encoding steps is halved but the difference between each phase encoding step
is doubled such that the remaining steps cover the same overall range. There are two
problems which arise due to this technique however. Firstly, the signal to noise ratio is
reduced and secondly aliasing can occur.

Aliasing, or wrap-around artefact

This occurs when the body part being imaged is bigger than the field of view selected.
The signal coming from the part of the body outside the field of view is misinterpreted
as coming from the other side of the image. Although this can theoretically occur in
both the phase encoding and the frequency encoding gradients, in practice the phase
encoding gradient is only applied to the field of view. Aliasing occurs in the frequency
encoding gradient if the returning signal is not sampled frequently enough. To avoid
this artefact the signal must be sampled at at least twice the wavelength of the highest
frequency. This (sampling twice per wavelength) is known as the Nyquist criteria. If the
signal is under-sampled then pixels from outside of the field of view are accidentally
mapped onto areas within the field of view. The classic example is the nose of a patient
appearing as an occipital mass.

Aliasing can be avoided by:

• Using a larger field of view

• Changing over the phase and frequency encoding gradients (if only a problem in
one direction)
• Applying a pre-saturation band to the area outside the field of view which
removes signal except from within the field of view
• Using software which corrects for the artefact.