MRI

Question 1

Describe the magnetic moment of atoms

Answer 1

A proton can be considered as a sphere of +ve charge, that is spinning on its axis
 Moving electric charge creates a magnetic moment µ
 The direction of µ is given by the right hand grip rule
 Hydrogen has the highest magnetic moment

Question 2

Why are hydrogen atoms the best for MRI?

Answer 2

 Hydrogen has the highest magnetic moment, hence provide a strong MRI signal. ( 1H1 contains only one
proton and there is no neutron to cancel or mask the proton spin value )
 Hydrogen is the most abundant atom in the body

Question 3

Which atoms have net spin?

Answer 3

Nuclei are made of protons and neutrons
 Both have spin (1/2)
 Pairs of spins tend to cancel
 So only atoms with an odd number of protons or neutrons have a net spin
 Good MRI nuclei are 1H, 13C, 19F, 23Na, 31P

Question 4

What happens when an external magnetic field is applied?

Answer 4

In an external magnetic field B0 ( >1 Tesla), protons will either line up its µ parallel or antiparallel to an external
magnetic field – no other orientation is possible
 The parallel state (blue sphere) has a lower energy (ground state) than antiparallel (orange) (excited state)
 This splitting in the spin energy level of a nucleus when placed in an external B field is known as Zeeman effect
 The energy difference between parallel and antiparallel state is Eflip
 The energy difference Eflip is proportional to the strength of external magnetic field B0

Question 5

What is Eflip?

Answer 5

The energy difference between parallel and antiparallel state is Eflip
 The energy difference Eflip is proportional to the strength of external magnetic field B

Eflip = Eantiparallel- Eparallel

Question 6

Describe the excess of protons in one state in comparison to the other

Answer 6

At room temperature, the number of protons in the lower state ≠ protons in higher state
 Slightly more than half of the protons are in lower energy state (i.e. aligned with B0
)
 For a 1 Tesla field, the relative excess is ~3 per million
 In a typical voxel in MRI, there are about 1021 protons and hence the excess protons in the lower energy state ~1015
 Hence, there is a net magnetisation (Mz) in the direction of external magnetic field

Question 7

Larger the Magnetic field B0 the …..

Answer 7

• Larger the difference in energy levels

* Larger the excess number aligned with field , i.e stronger Mz

Question 8

What happens when a RF frequency is applied?

Answer 8

If we deliver a photon of energy (hf = Eflip ), protons in the lower energy state may move to higher energy state

Question 9

What is precessional motion?

Answer 9

When placed in an external magnetic field, the proton also precess around the external magnetic field
 This precessional motion is due to the torque acting on proton due to external magnetic field
 This behaviour is similar to a spinning top, which wobbles due to force of gravity

Question 10

What is the Larmor frequency?

Answer 10

 The frequency of precessional motion is given by Larmor relationship
 Larmor frequency; f = (γ/2π) B0
 γ/2 π is known as gyromagnetic ratio, which is unique to each element
 Gyromagnetic ratio is expressed in MHz/T
 For hydrogen protons, the precessional frequency is 42.6MHz/T
 i.e. at 1T the protons precess at around 42 million times per second
 Irradiating the sample with an RF pulse of Larmor frequency will promotes
protons from lower energy state to higher energy state

Calculate the frequency of precession of 1H at 0.5, 1.5 and 3.0 Tesla?
frequency of precession @ 0.5 T = 42.58 x 0.5 = 21.29 MHz
frequency of precession @ 1.5 T = 42.58 x 1.5 = 63.87 MHz
frequency of precession @ 3 T = 42.58 x 3.0 = 127.MHz

Question 11

What is the gyromagnetic ratio?

Answer 11

γ/2 π is known as gyromagnetic ratio, which is unique to each element
 Gyromagnetic ratio is expressed in MHz/T
 For hydrogen protons, the precessional frequency is 42.6MHz/T
 i.e. at 1T the protons precess at around 42 million times per second

Question 12

What is T2 relaxation?

Answer 12

Decay of transverse magnetization
With the 900 RF pulse, a maximum net magnetization Mxy is generated in the transverse direction
 The transverse magnetic field Mxy induces a signal in the receiver coil (due to Faraday’s induction)
 This damped sinusoidal signal is known as Free Induction Decay (FID)
 After the RF signal applied, Mxy decays due to loss of coherence
 Loss of phase coherence is due to intrinsic micrcomagnetic inhomogeneities in the sample
 Due to the loss of phase coherence, the transverse magnetization Mxy decays
 T2 relaxation time is the time for the transverse magnetization to reduce to 37% of its initial value
 T2 relaxation is also known as spin-spin relaxation

Question 13

What is T2* decay

Answer 13

Extrinsic magnetic inhomogeneities (imperfect main magnetic field, presence of magnetic material
as contrast agents) add to the loss of phase coherence from intrinsic inhomogeneities and further
reduces the decay constant known as T2* decay

Question 14

What is spin-spin relaxation?

Answer 14

T2 relaxation

Question 15

Why T2 relaxation time is different for different tissues?

Answer 15

T2 relaxation is due to the loss of phase coherence as a result of intrinsic micro magnetic inhomogeneities,
whereby individual protons precess at different frequencies arising from the slight changes in local magnetic field
strength
 T2 values depend on the molecular structure of the sample
 Amorphous structures (CSF, edemous tissue, water) contain mobile molecules with fast and rapid motion. These
tissues do not support intrinsic magnetic field inhomogeneity, hence exhibit long T2
 Large macromolecules(bone) are non-moving and have large intrinsic field, hence very short T2
 T2 relaxation time doesn’t depend on field strength (very small dependence)

Question 16

What is T1 relaxation?

Answer 16

growth of longitudinal magnetization
Longitudinal magnetization Mz begins to recover simultaneously with transverse(Mxy) decay
 T1 relaxation time is the time needed for the recovery of 63% of Mz after 900 pulse
 T1 is also known as Spin-lattice relaxation time
 T1 value is affected by external magnetic field - higher B0 higher T1

Question 17

What is spin lattic relaxation time?

Answer 17

T1 relaxation time

Question 18

Why T1 relaxation time is different for different tissues?

Answer 18

T1 relaxation time depends on the rate of energy dissipation into the surrounding molecular lattice
 Molecules are tumbling through space with certain frequencies
 Energy transfer is most efficient, when this tumbling frequencies(vibrational frequencies) are comparable to
Larmour frequency
 Small molecule (water, CSF) have a broad distribution of motional frequencies with poor matching with Larmour
frequencies, hence shows long T1 values
 Medium sized molecules (proteins, fatty tissues) have a narrow distribution of tumbling frequencies and good
matching with Larmour frequencies. Hence short T1 values
 Large sized molecules tumble too slowly and almost no matching with Larmour frequency, hence very long T1

A- Large stationary molecules  Longest T1
C- Small aqueous molecules (water, CSF )  Long T1
B- Medium Viscous molecules(proteins, fatty tissues)  Short T1

Question 19

Compare T1, T2 & T2*

Answer 19

T1 of on the order of 5 to 10 times longer than T2
 Molecular size, motion etc. influence T1 and T2 relaxation
 For clinical MRI application, most tissues of interest are intermediate to small sized molecule
 In these ranges, tissue with longer T1 usually have a longer T2 and those with shorter T1 have shorter T2
 In summary T1>T2>T2*

Question 20

Why pulse sequence ?

Answer 20

Molecular size, motion etc. influence T1 and T2 relaxation
 T1, T2, T2* decay constants and proton density are fundamental properties of tissues
 T1 is longer for higher magnetic field strengths, T2 is almost independent on magnetic field
 By emphasising the differences of T1 and T2 relaxation time constants and proton density of tissues, the contrast
of MRI image can be changed
 This can be done by using a pulse sequence
 A pulse sequence defines the timing and nature of RF signal that generates the transverse magnetization
 e.g. 900 – 1800
- 900
 An MRI pulse sequence dramatically impacts the appearance of an MRI image

Question 21

What is Time of Repetition (TR)

Answer 21

Acquisition of an MRI image relies on the repetition of sequence of events (e.g. repetition of 900
-1800 sequence)
 Time of repetition (TR) is the period between RF excitation pulse
 During the TR interval, T2 decay and T1 recovery occur in the tissue
 TR values range between few milliseconds to very long (10,000ms), depending on the type of sequence employed

Question 22

What is Time of Echo (TE)?

Answer 22

900 RF excitation creates transverse magnetization and FID signal
 To separate RF energy deposition and returning signal, an echo is induced
 Echo is often induced with a 1800
inversion pulse
 The time of echo (TE) is the time between the excitation pulse and the appearance of the peak amplitude of an
induced echo
 Echo is induced by applying a 1800 RF inversion pulse at a time TE/2
 Echo can also be induced by gradient polarity reversal

Question 23

Three major pulse sequences are used for MRI image acquisition

Answer 23

1. Spin Echo (SE)
2. Inversion Recovery (IR)
3. Gradient Echo (GE)

Question 24

What is Spin Echo (SE) pulse?

Answer 24

Spin Echo imaging technique has the advantage that it is insensitive to static inhomogeneity of the magnet and
inhomogeneity caused by magnetic susceptibility(ability of a substance to become magnetized) of patient tissue
 Most commonly used pulse sequence
 Spin Echo sequence exists in many forms- multi echo pulse sequence, fast spin echo (FSE), echo planar imaging,
gradient and spin echo (GRASE) – all are essentially spin echo sequence
 As the name implies, this technique involves the generation of an echo signal
 Spin Echo pulse sequence consists of 900 RF pulse to excite the magnetization and followed by 1800 RF pulse to
refocus the spins to generate signal echo known as spin echo (SE)

Question 25

Steps in spin echo pulse

Answer 25

First, a 900 RF pulse is applied
• The 900 RF pulse converts Mz into Mxy (i.e. longitudinal to transverse magnetization)
• Soon after 900 RF pulse the transverse magnetization decays due to loss of coherence of protons
described by T2* (extrinsic inhomogeneities)
• This generates a FID signal
 Second, a 1800 RF pulse is applied at TE/2
• The 1800 RF pulse inverts the spin system and cancels the extrinsic inhomogeneities , hence the
dephasing effect
• As a result, recovery of transverse magnetization occurs, which generates an echo signal at time TE

Question 26

Role of inversion pulse to cancel the effects of extrinsic inhomogeneities

Answer 26

 An external static magnetic field of 1.5T have inhomogeneities of 1ppm, then the T2* will be 2.5ms
 This is typically much smaller than the relaxation time of most tissues and will mask any attempt to generate
images based on T2 weighting
 By applying 1800 pulse the effects of extrinsic inhomogeneities will be cancelled
 i.e. No T2* occurs with Spin Echo sequence due to 1800
refocusing pulse (more robust against susceptibility
artefacts)

Question 27

Spin Echo contrast weighting

Answer 27

Contrast is proportional to the difference in signal intensity between adjacent pixels in an image
 The signal intensity produced by an MR system for a specific tissue using a SE sequence is given by
S=rH
[1-e
TR/T1]e-TE/T2
 rH
is the proton density, T1 and T2 are physical properties of tissue and TR and TE are pulse sequence timing
parameters
 By changing the pulse sequence parameters TR and TE, the contrast dependence can be weighted towards T1,
proton density or T2 characteristics of the tissue

Question 28

T1 weighting

Answer 28

 A T1 weighted SE sequence produce contrast mainly based on the T1 characteristic of tissue with de-emphasis
of T2 and proton density contribution to the signal
 Achieved by using a short TR to maximise the difference in longitudinal magnetization recovery and short TE
to minimise T2 decay
 Typically TR = 400-600 ms and TE = 3-30 ms
 T1 weighted image produces good contrast between soft tissue types (because different tissues have different
T1 values)
 Remember - T1 values; Fat (260 ms) White matter (780 ms) Gray matter (900 ms) and CSF (2400 ms)
 Fat with short T1 has a large signal because of the greater recovery of Mz
 CSF with long T1 has a low signal
 i.e. Fat appear white and CSF appear dark

Question 29

Proton density(PD) weighting

Answer 29

 Proton density contrast weighting relies mainly on differences in the number of magnetized protons per unit
volume of tissue
 At equilibrium, tissues with large proton density (lipids, fat, CSF) have a large MZ
compared to other soft tissue
 Achieved by reducing the contribution of T1 recovery and T2 decay
 T1 differences are reduced by selecting a long TR
 T2 differences of the tissue are reduced by selecting a short TE
 Signal strength (contrast) depends on the number of protons CSF> fat> gray matter> white matter
 Typically, TR = 2000-4000 ms and TE = 3-30 ms

Question 30

T2 weighting

Answer 30

 T2 weighted image demonstrates good contrast between normal tissue and pathology (many pathologies have
elevated T2 values due to increased water content)
 Reduce T1 differences in tissue with long TR
 Emphasize T2 difference with long TE
 Typically, TR=2000-4000ms and TE=60-150ms
 T2 values: Fat (80ms), White Matter (90 ms) and Gray matter (100) CSF (160ms)
 i.e. CSF produces a maximum signal (appear white) while fat appears dark [opposite to T1 weighted image]

Question 31

T1, PD and T2 weighting image acquisition

Answer 31

T1 contrast
TR(ms) - 400-600 (SHORT)
TE(ms) - 3-30 (SHORT)
Fat-bright
WM
GM
CSF -dark

PD contrast
TR(ms) - 2000-4000 (LONG)
TE(ms) - 3-30 (SHORT)
CSF bright
GM
WM
Fat- dark

T2 contrast
TR(ms) - 2000-4000 (LONG)
TE(ms) - 60-150 (LONG)
CSF bright
GM
WM
Fat- dark

Question 32

Gradient Echo Pulse Sequence

Answer 32

 Another pulse sequence used in MRI
 GE technique uses a magnetic field gradient applied in one direction and then reversed to induce an echo (instead
of 1800 pulse)
 The major variable determining tissue contrast in GE is the initial flip angle (< 600
is commonly used )
 GE technique has great versatility- A variety of contrasts can be produced while imaging rapidly
 Deposition of RF energy in the patient is lower since the 1800 RF pulses are not used (less heating of patient
tissues)
 3D or volume imaging can be accomplished
 But, static inhomogeneity of the magnet and inhomogeneity caused by magnetic susceptibility of patient tissue
are not corrected by GE (echo in the same direction as FID)

Question 33

Inversion Recovery Pulse Sequence

Answer 33

 Another pulse sequence used in MRI
 IR emphasize T1 relaxation time of the tissues by extending the amplitude of longitudinal recovery by a factor of 2
 First, 1800 RF excitation pulse is applied
• This inverts the longitudinal magnetization vector through 1800 (i.e. Mz becomes –Mz)
• The magnetization vector begins to relax back to Mz (2 time recovery time )
 Next, a 900 RF pulse is applied after a time from 1800 pulse known as the TI (time of inversion)
 Next, a second 1800 pulse is applied at TE/2, which refocuses the transverse magnetization and generates an echo
at TE
 The signal strength is chiefly a function of the T1 characteristics of the tissues

Question 34

Advantages of inversion recovery pulse sequence

Answer 34

 Inversion recovery pulse sequence are useful for suppression of selected tissues (orbital fat, fatty tumours, CSF)
 IR creates heavily T1 weighted images without dominant contribution from fat (e.g. brain, liver)
 The 1800
inverting pulse can produce a large contrast between fat and water
 Variations of IR sequence
• STIR (Short TI Inversion recovery) sequence – to suppress the signal from fat
• FLAIR (Fluid Attenuation Inversion Recovery) sequence – to suppress the signal from water

Question 35

What are gradient coils?

Answer 35

Spatial localization is essential to determine the localization of sample volume
 Achieved by superimposing linear magnetic field variations on the main magnetic field
 Typical gradient field strength ranges from 1-50 mT/m
 This generates corresponding position dependent variations in precessional frequency of protons
 Gradient coils are also used to apply reversal pulses in some imaging technique (Gradient Echo sequence)

Question 36

Slice Selection Gradient (SSG)

Answer 36

Localize along z-axis
 Slice Select Gradient (SSG) applied along the long axis of the body
 The proton processional frequencies are incrementally increased or decreased dependant on their distance from
the null
 Now, a selective narrow band of RF pulse is used to excite protons from a specific location
 e.g. 63.87MHz RF pulse will excite only the protons at null position

Question 37

What is Frequency Encode Gradient (FEG)

Answer 37

Localize along z-axis
 Slice Select Gradient (SSG) applied along the long axis of the body
 The proton processional frequencies are incrementally increased or decreased dependant on their distance from
the null
 Now, a selective narrow band of RF pulse is used to excite protons from a specific location
 e.g. 63.87MHz RF pulse will excite only the protons at null position

Question 38

What is Phase Encode Gradient (PEG)?

Answer 38

Position of the protons in the third orthogonal dimension is determined with a PEG
 i.e. to localize along y-axis

Question 39

What is “k- space” data acquisition?

Answer 39

 MR data are initially stored in the k space matrix
 “k -space” matrix is a frequency domain repository
 The spatial frequency signals acquired during the evolution and decay of echo is stored in k space
 “k -space” describes a two dimensional matrix of positive and negative spatial frequency values

Question 40

Image reconstruction

Answer 40

 Data are deposited in the k space matrix determined by FEG and PEG
 A process known as “inverse two dimensional Fourier Transform” converts data into a visible image (convert
frequency domain to space domain)
 Final image is scaled and adjusted to represent the proton density, T1 and T2 characteristic of the tissue using a
grayscale range, where each pixel represent a voxel

Question 41

Functional MRI (fMRI)

Answer 41

 MRI can also be used to obtain functional information
 Blood flow increases more than usual to active regions of the brain
 This results in a local reduction of deoxyhemoglobin
 Deoxyhemoglobin (contains iron) is a paramagnetic agent, it alters T2* weighted MRI signal
Oxyhemoglobin also contains iron, but bound to oxygen.
• In case of deoxyhemoglobin, there are four free electron which causes inhomogeneities in the external
magnetic field and contribute to T2* effect

Question 42

MRI contrast agents

Answer 42

 MRI is highly sensitive but not very specific
 Contrast agent improves specificity
 Contrast agents also increase signal to noise ratio, SNR (i.e. better image)
 MRI contrast agents alter the relaxation times
 Most commonly used contrast agent is Gadolinium or “gado” or “gad”(Gd )
 7 unpaired electrons
 Large paramagnetic susceptibility
 Shorten T1- hence enhance signal on T1 weighted image
 Enhance signals from blood vessels, brain tumour imaging etc.
 SPIO (Super paramagnetic Iron Oxide)
 Reduces T2- lower intensity on T2 weighted image
 Liver/spleen image
 lower normal tissue signal- pathological tissue enhanced

Question 43

MRI Artefacts

Answer 43

Susceptibility Artefacts
• Due to the large change in magnetic susceptibility
across the field of view
• e.g. metal implants, tissue air interface

RF Artefacts
• RF pulse and precessional frequencies of MRI occupy the same frequencies of
common RF sources such as TV, FM radio, computers, fluorescent light etc.
• Stray RF signals that propagate to the MRI antenna

Motion/Flow Artefacts
• Most common and noticeable artefact due to moving anatomic structures (eyes, heart, lungs etc) and flow
(blood, CSF)
• Increases with long MRI pulse sequences
• Motion artefacts predominantly propagate in phase encoding direction (FEG direction is less affected)
• Therefore, phase encoding often is chosen along the shortest dimension
• Several techniques can be used to compensate for motion-related artefacts

Chemical Shift
• Proton precessional frequency in water and fat are slightly
different- 3ppm difference
• Precessional frequency in fat is lower
• This difference in frequency is translated onto a different
position in the image

k space error
• Error in k space encoding affect all of the reconstructed
images
• This causes artefactual superimposition of wave pattern
across the FOV

Question 44

MRI room – Faraday Cage

Answer 44

 MRI signal is obtained in the form of RF signal
The room must be shielded against any outside interference (FM signal, monitoring equipment)
A Faraday cage is used
Faraday cage is an enclosure formed by a conductive material (e.g. Cu) or by a mesh of such material
Faraday cage distribute charge (radiation) around the cage’s exterior and hence cancels out the
field’s effect in the cage’s interior
Faradays cage isolate the MRI scanner from outside interference
Also, stops RF frequencies produced by the scanner from interfering with sensitive medical
equipment outside the room

Question 45

Main magnet

Answer 45

 Superconducting electromagnet is often used to produce the main magnetic field
 The coil of the wire is kept at a temperature of 4.2K by immersing in liquid helium(~ 1700 litres)
 This is surrounded by two chambers filled with liquid nitrogen (15K)
 Field strengths from 0.5T to 9T (1.5T most commonly used)
 Quite expensive and difficult to maintain

Question 46

Safety and biohazard

Answer 46

 Although no ionising radiation is used, many important bio effects and safety issues to be considered
 Strong magnetic field- has a fringe field extend to several meters
 Controlled access to areas where magnetic field is >0.5mT
 RF energy
 Time varying magnetic field gradient
 Cryogenic liquid
 Confined imaging device (claustrophobia)
 Very noisy operation(gradient coil activation and deactivation)
 Metallic implants, prostheses, heart valves, pacemakers
 Non metallic implants can also lead to significant heating under rapidly changing gradient field
 Ferromagnetic materials brought into the room (e.g. an IV pole) can become deadly projectiles

Question 47

emergency quench

Answer 47

 Remember- even when you are not scanning the magnet is not OFF
 A quench should only be performed by authorised personnel with proper training in dire emergency
 In extreme emergencies, superconducting magnet can be turned off by a controlled “quench” procedure
 The quench procedure subjects the magnet to a 2600
temperature difference in a short period of time
 If performed quickly, major damage to the magnet can occur
 Sudden loss of superconductivity may result in explosive boiling of liquid helium and jeopardise the safety of
those in the room and adjacent areas
 In the event of magnet quench, the patient and all other personnel must be immediately evacuated from the
examination room