/
G16.4427 Practical MRI 1 G16.4427 Practical MRI 1

G16.4427 Practical MRI 1 - PowerPoint Presentation

mitsue-stanley
mitsue-stanley . @mitsue-stanley
Follow
384 views
Uploaded On 2016-06-30

G16.4427 Practical MRI 1 - PPT Presentation

Advanced pulse sequences Signal Formula for SE M xy negligible TR gtgt T2 or spoiler gradient 90 180 M zA short pulse no T 1 relaxation between A and B or C and D ID: 384026

pulse gradient phase echo gradient pulse echo phase image diffusion sequences mri fat epi images handbook water echoes bernstein

Share:

Link:

Embed:

Download Presentation from below link

Download Presentation The PPT/PDF document "G16.4427 Practical MRI 1" is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.


Presentation Transcript

Slide1

G16.4427 Practical MRI 1

Advanced pulse sequencesSlide2

Signal Formula for SE

M

xy

negligible

(TR >> T2, or

spoiler gradient)

= 90°

= 180°

M

zA

short pulse (no T

1

relaxation

between A and B, or C and D)

Bernstein et al

. (2004) Handbook

of MRI

Pulse SequencesSlide3

Multi-Echo SE

The transverse magnetization can be repeatedly refocused into subsequent

SEs

by playing additional RF refocusing pulse

The series of echoes is called an echo train

Each echo number fits its own independent k-spaceThe length of the echo train is limited by T2

decayIn most cases we are interested in 2 echoes (an early and a late one).

Question: if TR is long, what contrast will have the 2 resulting images?Slide4

Multi-Echo SE

The transverse magnetization can be repeatedly refocused into subsequent

SEs

by playing additional RF refocusing pulse

The series of echoes is called an echo train

Each echo number fits its own independent k-spaceThe length of the echo train is limited by T2

decayIn most cases we are interested in 2 echoes (an early and a late one).

 if TR is long, the two images will be PD- and T2-weighted, respectivelySlide5

Example of Dual-Echo SE Acquisition

Proton density-weighted

TE/TR = 17/2200 ms

T

2

-weighted

TE/TR = 80/2200 msSlide6

Dual-Echo SE

Bernstein et al

. (2004) Handbook

of MRI

Pulse SequencesSlide7

T2-Mapping

It is a common application of acquiring longer echo trains (otherwise more than two echoes per TR are rarely acquired in MRI)

In theory we can acquire long echo train of

SEs

and fit the signal intensity at each pixel to calculate T

2In practice there are systematic errors that make it difficult to fit a monoexponential decay curve

Variable flip angle across slice profileStimulated echoes can introduce unwanted T

1-weighting variations into the echo-train signalsIf magnitude reconstruction is used, the noise floor has nonzero mean leading to incorrectly larger T2 valuesSlide8

Paper DiscussionSlide9

Advanced pulse sequencesSlide10

Echo Planar Imaging (EPI)

EPI is one of the fastest MRI pulse sequences

2D image in few tens of milliseconds

Allowed developing challenging MR applications

Diffusion, perfusion, cardiac imaging, etc.

EPI uses a gradient-echo trainTypical to produce ~100 gradient echoes to produce a low-resolution image from a single RF excitationMore prone to a variety of artifactsGhosting along phase-encoded directionSlide11

Ghosting Artifacts

Not caused by motion, but by eddy currents, imperfect gradients, field non-uniformities, or a mismatch between the timing of the even and odd echoes

Which results in

mis

-registration of alternating lines of k-space

Phase errors may result

from the multiple positive

a

nd negative passes through k-spaceGhost artifacts in the

phase directionSlide12

GRE vs. EPI

In a simple GRE pulse sequence the transverse magnetization decays as

M

xy

(t

) = Mxy(0)e-t/T2*Half lifetime of

Mxy is T

2*ln2A very small fraction of the lifetime is actually used for data acquisition in GREEPI maximally uses the transverse magnetization without additional RF excitationsBipolar oscillating readout gradient produces a series of echoes, each individually phase-encodedSlide13

GRE and EPI

N

etl

= echo train length = number of echoes following RF excitation

t

esp

= echo spacing (typically echoes are evenly spaced)

A series of echoes

is produced before

M

xy decays away due

to T2

* relaxation

Bernstein et al. (2004) Handbook

of MRI Pulse SequencesSlide14

EPI Readout Gradient

Starts with a

prephasing

gradient that position the

k

-space trajectory to k

x,min

followed by a series of readout gradient lobes with alternating polarity. Question: what is the area of the prephasing gradient lobe Gx,p?

Bernstein et al

. (2004) Handbook

of MRI

Pulse SequencesSlide15

EPI Readout Gradient

Starts with a

prephasing

gradient that position the

k

-space trajectory to k

x,min

followed by a series of readout gradient lobes with alternating polarity. Answer: Half the area of the first readout gradient area.

The second half of

each lobe serves as

prephasing

for the

following gradient

(that is why the polaritymust alternate)

Bernstein et al

. (2004) Handbook of MRI Pulse SequencesSlide16

EPI Phase-Encoding Gradient

Constant phase-encoding gradient throughout the entire readout echo train (

k

y

varies linearly with time)

Gridding is needed before reconstruction

A series of blips with the same polarity and identical area, each played before the acquisition of an echo

Bernstein et al

. (2004) Handbook

of MRI

Pulse SequencesSlide17

Gradient-Echo EPI

Bernstein et al

. (2004) Handbook

of MRI

Pulse SequencesSlide18

Spin-Echo EPI

What is different compared to GRE-EPI?Slide19

Phase Contrast (PC) Imaging

A method to image moving magnetization by applying flow-encoding gradients

Image flow in blood vessels and CSF, track motion

The flow-encoding gradients translates velocity into the phase of the image

Bipolar gradient, as it produces linear proportionality

The axis of the gradient determines the direction of flow sensitivityNormally applied to only one axis at a timeTypically performed with GRE pulse

sequences, adding phase-encoding

gradients. Why?Slide20

Typical PC Pulse Sequence

Typically a bipolar gradient is added to only one of the three logical axes at a time

Toggling of the bipolar gradient (dotted lines) varies the gradient first moment and introduces flow sensitivity along that axis

Bernstein et al

. (2004) Handbook

of MRI

Pulse SequencesSlide21

PC Acquisition and Reconstruction

Two complete sets of image are acquired varying only the 1

st

moment of the bipolar gradient

The amount of such operator-selected variation determines the amount of velocity encoding

The phases of the two images are subtracted on a pixel-by-pixel basis in image domainAllows to quantify flow direction, flow velocity and volume flow ratePhase-difference or complex-difference reconstruction methods are in common useSlide22

Diffusion Imaging

In the presence of a magnetic field gradient, diffusion of water molecules causes a phase dispersion of the transverse magnetization

The degree of signal loss depends on tissue type, structure, physical and physiological state

Diffusion imaging is a family of techniques

E.g. DWI, DTI, DKI

All diffusion pulse sequences contain diffusion-weighting gradientsDWI typically employs a single b

-value, other quantitative mapping methods at least 2 b-valuesSlide23

Diffusion-Weighting Gradients

Typically consist of two lobes with equal area

Amplitude is the maximum allowed

Pulse width is larger than most imaging gradients

When used, water diffusion can cause an attenuation in proton MRI signals

Degree of attenuation depends on the product between the diffusion coefficient D and the b-valueb

-value is analogous to TE in T2-weighted sequencesIncreasing gradient amplitude, separation of its lobes, or pulse width of each lobe results in a higher

b-value. How does diffusion-weighting change consequently?Slide24

Diffusion Weighting in GRE and SE

Spin Echo

Gradient Echo

Bernstein et al

. (2004) Handbook

of MRI

Pulse SequencesSlide25

Single-Shot Spin-Echo EPI

The most prevalent sequence due to high acquisition speed (e.g. < 100 ms per image)

A pair of identical gradient lobes on either side of the refocusing pulse

Gradient direction can be controlled by varying its vector components along the 3 axes

To minimize TE the max amplitude is used to achieve the desired

b-valueWhat is another way to minimize TE?Slide26

Single-Shot Spin-Echo EPI

The most prevalent sequence due to high acquisition speed (e.g. < 100 ms per image)

A pair of identical gradient lobes on either side of the refocusing pulse

Gradient direction can be controlled by varying its vector components along the 3 axes

To minimize TE the max amplitude is used to achieve the desired

b-valueMaximizing SR also reduces TE, but can cause peripheral nerve stimulationSlide27

Diffusion-Weighted Single-Shot SE

Bernstein et al

. (2004) Handbook

of MRI

Pulse SequencesSlide28

Diffusion-Weighted (DW) Imaging

In the presence of a gradient, molecular diffusion attenuates the MRI signal exponentially:

Tissue with fast diffusion experiences more signal loss

low intensity in the DW image

To remove the patient-orientation dependence, 3 DW images can be obtained with a DW gradient applied along the three orthogonal directionsIf same

b-value then isotropic DW image

(S and S

0

are the voxel signal intensity with and without diffusion)Slide29

Example: White Matter InfarctSlide30

Quantitative Apparent Diffusion Coefficient (ADC) Mapping

A series of DW images are acquired with multiple

b

-values:

A linear fit between ln(S

0/Si) and

bi is performed on a pixel-by-pixel basis to find

DThe contrast of the ADC map is inverted compared to a DW imageTo keep a constant TE in all DW images, b-values are typically changed by varying the diffusion-gradient amplitude instead of its durationContribution from imaging gradients should be included in b-value calculation to avoid overestimating ADC

…Slide31

DW Image vs. ADC MapSlide32

Fat Suppression

Because of its short T

1

, the bright appearance of fat is a problem for T

1

-weighted images with short TR and short TEFat is a main contributor to chemical shift artifactsThere are several methods for fat suppressionSpectrally selective RF pulsesWhat are the drawbacks?Slide33

Fat Suppression

Because of its short T

1

, the bright appearance of fat is a problem for T

1

-weighted images with short TR and short TEFat is a main contributor to chemical shift artifactsThere are several methods for fat suppressionSpectrally selective RF pulsesB

1 and B0 inhomogeneities, not good at low field strengths

Short TI recovery (STIR)B1 inhomogeneities, long scan, signal from other tissuesSlide34

Two-Point Dixon Pulse Sequence

In conventional spin echo

Δ

= 0

Question:

what happens if

Δ

≠ 0

Bernstein et al

. (2004) Handbook

of MRI Pulse SequencesSlide35

Two-Point Dixon Pulse Sequence

In conventional spin echo

Δ

= 0

If

Δ

≠ 0, spins with different chemical shifts will be out of phase at

k

x = 0 (unless phase shift happens to be a multiple of 2π)

If the 180° pulse is delayed or advanced by Δ/2, the RF spin echo is delayed or advanced by

Δ

relative to where k

x = 0

Consider a voxel with a water and a fat spin having a frequency difference f

cs and assume there are no B0

inhomogeneitiesIn the 2-point Dixon technique, we acquire two SE images:

Δ = 0 and normal acquisitionFat and water in-phase at k

x = 0 (in-phase image)

Δ = 1/(2fcs) and RF pulse advanced or delayed by Δ/2 = 1/(4f

cs)

Fat and water 180° out-of-phase at k

x = 0 (out-of-phase image)

ϕ = 2π

fcsΔ at

kx

= 0 Slide36

Two-Point Dixon Technique

Because image contrast is heavily determined by the peak signal amplitude that occurs

at

k

x

= 0, the resulting complex images are approximately given by:

I

0 = W + F I

1 = W – F

Separate images of the water and fat magnetization can be reconstructed from:

W

= (1/2) (

I

0

+

I

1

) F = (1/2) (I0

- I1)

The water image W

can be used as a fat suppressed image, whereas W and F separately provide information about the relative water and fat content of tissuesSlide37

Limits of Two-Point Dixon

The 2-point Dixon technique assumes perfect

B

0

homogeneity which is almost never true

Due to ΔB0 fat and water have accumulated an additional phase shift Δ

ϕ = γ(ΔB

0)Δ at kx = 0 in I1Although fat and water spins in any given voxel are still anti-parallel in the opposed-phase image, they might no longer be parallel or anti-parallel to the fat and water spins in the in-phase image

(example of 90° phase shift caused by

B

0

inhomogeneities)

Question:

what is wrong with the combined images?Slide38

Example: Healthy Liver

AJR April 2010, vol. 194(4),

p

. 964-971Slide39

Example: Fatty Liver

AJR April 2010, vol. 194(4),

p

. 964-971Slide40

Any questions?Slide41

See you next Thursday!