G16.4427 Practical MRI 1

G16.4427 Practical MRI 1 G16.4427 Practical MRI 1 - Start

Added : 2016-04-09 Views :48K

Download Presentation

G16.4427 Practical MRI 1




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.



Presentations text content in G16.4427 Practical MRI 1

Slide1

G16.4427 Practical MRI 1

Pulse Design for Parallel MR Transmission

Slide2

Outline

K-space analysis of small tip-angle excitation

RF shimming and Parallel MR Transmission

B

1

mapping

Accelerated MR excitations

Homogeneity and SAR minimization

Slide3

K-Space Interpretation of Small Tip Angle Excitation

New point of view for analyzing selective excitationSimilar approach as that of data acquisition and reconstructionStrictly valid only for small tip angles excitations but holds also for flip angles ~ 90°

Bloch equation in the rotating frame

(ignoring T

1

and T

2

)

Slide4

Excitation K-Space

(parametrically describes a path through the spatial frequency space)

Parametric description of the unit-weight trajectory (

sampling structure)

Spatial-frequency weighting of the

k

-space trajectory

Slide5

Application to Slice-Selective Excitation

Conventional slice-selective excitation pulse sequence

k

-space interpretation

K-space is scanned linearly as the RF field is applied

The location in

k

-space at time

t

is the integral of the remaining gradient waveform

Origin reached when the remaining waveform integrates to zero

RF weighting is symmetric with respect to the origin

Slice profile (Fourier transform of RF weighting) is in phase

The role of the refocusing lobe is to shift the

k

-space origin back in the middle of the RF excitation

Slide6

Multiple Coil Excitations

a

1,

φ1

a2, φ2

a8, φ8

a7, φ7

a6, φ6

a5, φ5

a3, φ3

a4, φ4

RF

RF

RF

RF

RF

RF

RF

RF

RF

a

2

(t)

,

φ

2

(t)

a

3

(t)

,

φ

3

(t)

a

4

(t)

,

φ

4

(t)

a

1

(t)

,

φ

1

(t)

a

7

(t)

,

φ

7

(

t)

a

6

(t)

,

φ

6

(t)

a

5

(t

)

,

φ

5

(t)

a

8

(t)

,

φ

8

(t)

RF shimming

Distinct but time-constant amplitudes and phases for each element

Common gradient and RF waveform

Parallel Transmission

Distinct and time-varying amplitudes and phases for each element

Common gradient waveform but distinct RF waveform

Slide7

Parallel RF Transmission

Parallel transmission may be used to correct RF inhomogeneities, control SAR, tailor excitationsRequires calibration of coil array excitation patterns, and operates in close analogy to parallel reception

Slide8

Small Flip Angle Excitation

Homogeneous volume coil

excitation

k

-space sampling trajectory

(controlled by the switching gradients)

spatial-frequency weighting

(proportional to the coil driving current)

Transmit coil array

(

M

xy

is obtained by multiplying the profile by

iγM

0)

B

1 spatialweighting

Question: what is the B1 spatial weighting?

Effective spatial weighting, to account for coupling-induced

intercoil

correlations

Slide9

Illustration of Parallel Transmission

1 RF pulse

+ the gradient pulse

Slide10

Illustration of Parallel Transmission

1 RF pulse

+ the gradient pulse

+ B

1

weighting

Slide11

Illustration of Parallel Transmission

L (coils) RF pulses

+ the gradient pulse

+ B

1

weighting

Slide12

Example: 2D Selective Excitation

Slide13

Example: 3D Selective Excitation

Slide14

Outline

K-space analysis of small tip-angle excitation

RF shimming and Parallel MR Transmission

B

1

mapping

Accelerated MR excitations

Homogeneity and SAR minimization

Slide15

B1 Mapping

Accurate transmit RF field (B

1

+

) or flip angle maps are needed for many MR applications.

Examples?

Correct the results of quantitative methods

Validate theoretical models for EM calculations

Testing MR compatibility of implanted objects

Compensate for B

1

inhomogeneities

Image-based RF field measurements are needed for

in-vivo

applications

Several B

1

mapping techniques exists, but further improvements (time efficiency, anatomical coverage, accuracy) are needed to use them in the routine practice and for parallel transmission

Slide16

Multi-Point Intensity Method

Non-selective RF pulse (long TR) and FIDSignal is largely independent from T1 and T2S ∝sin(α)Step through transmit voltage until the first signal maximum is found (i.e. α = 90°)Other pulse amplitudes would then be set relative to this calibration pulseFor GRE:

Slide17

Double Angle Method (DAM)

Collect two scans, one of which uses twice the RF amplitude of the other.

image value at

pixel

j

object magnetization

at voxel

j

unknown actual

flip angle

error

“Double angle formula”

Slide18

Double Angle Method (DAM)

Collect two scans, one of which uses twice the RF amplitude of the other.

image value at

pixel

j

object magnetization

at voxel

j

unknown actual

flip angle

error

Inefficient method (TR

5T

1

required)

Performs poorly in regions of low signal

ambiguities

if

is

too large, sensitive to noise if

is too small

“Double angle formula”

Slide19

Phase-Based Method

Exploits the fact that rotations do not commuteFinal Mxy differs by a phase that depends on the magnitude of the flip angle α

α

x

α

y

α

-

x

α

-

y

α

x

 αy

αy

 αx

 α-y

 α-x

 αy

 αx

Slide20

Question:

What are pros and cons of the phase-based method?

Slide21

Phase-Based Method

Exploits the fact that rotations do not commuteFinal Mxy differs by a phase that depends on the magnitude of the flip angle α

Works better for larger α and shorter pulses ( SAR limitation)Only for 3D and sensitive to motion/flow

α

x

α

y

α

-

x

α

-

y

α

x

 αy

αy

 αx

 α-y

 α-x

 αy

 αx

Slide22

Actual Flip Angle Imaging (AFI)

Two identical RF pulses followed by two delays of different duration (TR1 < TR2 < T1)

Assumption: at the end of both TR1 and TR2 the transverse magnetization is completely spoiled (need RF spoiling with dummy repetition to reach steady state)

Before each excitation pulse:

The observed signals are:

Their ratio is:

Slide23

Any questions?

Slide24

Outline

K-space analysis of small tip-angle excitation

RF shimming and Parallel MR Transmission

B

1

mapping

Accelerated MR excitations

Homogeneity and SAR minimization

Slide25

Parallel Transmit For 2D EPI Excitation

Complex-valued excitation profile

Periodic excitation pattern associated with the RF pulse of the

l

th

transmit coil

In the case of a 2D EPI excitation trajectory, let’s define:

If the sampling interval in excitation

k

-space is sufficiently small, then

Δx

= 1/Δ

k

x

is big enough that all the aliasing lobes are outside the FOV:

If we

undersample

(i.e. use a larger sampling interval) in excitation

k

-space, then M lobes will alias inside the FOV:

Slide26

Accelerated Parallel MR Excitations

(in the case of the EPI excitation trajectory we can treat each position separately)

To design our parallel transmit pulse design we need to find the periodic excitation patterns for each transmit coil such that:

Question:

how would the equation above change for an accelerated RF excitation?

Slide27

Accelerated Parallel MR Excitations

(in the case of the EPI excitation trajectory we can treat each position separately)

To design our parallel transmit pulse design we need to find the periodic excitation patterns for each transmit coil such that:

We can exploit the extra degrees of freedom to under sample the excitation by a factor

M

:

Slide28

Outline

K-space analysis of small tip-angle excitation

RF shimming and Parallel MR Transmission

B

1

mapping

Accelerated MR excitations

Homogeneity and SAR minimization

Slide29

SAR and RF Homogeneity

SAR management and RF homogeneity are critical issues at high magnetic field strengths

SAR is a potentially elevated safety concern

B

1

focusing compromises

the underlying SNR increase

Slide30

RF Power Deposition in Multiple Coil Excitations

electric field covariance matrix

RF energy dissipated in

noise covariance matrix

Net Electric Field

EPI excitation trajectory

Small tip angle

Global SAR

Image-domain global SAR

unit current electric field

RF excitation patterns

Slide31

Pulse Design for SAR Reduction

weighting

Homogeneous excitation

with minimum SAR

target excitation

profile at

Minimum global SAR

Optimal excitation patterns

for Parallel Transmission

Time-independent RF shimming

Optimal modulation coefficients

shared excitation profile

optimal modulation

Slide32

Minimum SAR with Parallel Tx

SAR =

3.3

20 coils

SAR =

5.5

12 coils

SAR =

7.9

8 coils

Parallel Transmission - 7 Tesla - No Acceleration

SAR =

1

Ultimate

Basis Set

Slide33

SAR vs. Profile Homogeneity

B

o

= 7T

Slide34

Calibrating the Phi Matrix

2

3

4

1

3

4

1

2

Slide35

Calibrating the Phi Matrix

2

4

1

3

3

1

2

4

Slide36

Calibrating the Phi Matrix

4

1

3

2

3

4

1

2

Slide37

How Many Measurements?

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Experiment

Channel

1

2

3

4

4 real

diagonal elements6 complex off-diagonal elements

16 variables to estimate

16 (# of channels × # of channels) measurements required

Question:

Why 16 measurements?

Slide38

Simulation Results

1

2

3

4

Φ

calibrated using

simulated E fields

Φ

calibrated using

power measurements

Identical !

Measured and predicted power

Simulation set up

Slide39

Experimental Set Up (7 Tesla)

Directional couplers

RF switch

National Instrument Dual 16x1 MUX

Knee setup,

8-channel parallel

Tx

stripline

coil

Power meter

Rhodes & Schwarz

NRP

-Z11

Computer that automates the measurement

Slide40

In-Vivo Results

40-miliseconds

Φ

calibrated at 60V

40-miliseconds

Φ

calibrated

at

120V

measured

power

predicted

power

Slide41

Uses of Phi Matrix Calibration

Prediction and real-time monitoring of global SAR

Prediction and real-time monitoring of individual channel FWD and RFL power

Real-time detection of

Tx

chain hardware failures

Optimization of RF pulse design for RF shimming and parallel transmission

Slide42

Maximum Efficiency RF Shimming

Array transmit efficiency metric:

Average B

1

+

strength squared:

Total power deposition

# of spatial locations

It can be treated as a generalized

eigenvalue

problem:

Largest

eigenvalue

= maximum transmit efficiency

η

max

Corresponding eigenvector =

w

max

for maximum efficiency RF shimming

Maximum Efficiency RF Shimming

Find

w

that maximize

η

Slide43

Experiment: Hip Imaging at 7 T

Flip Angle Maps

90

0

π

Excitation: 4

ch

Tx

/Rx loop coils

Receive: 10

ch Tx/Rx (5 loop/stripline modules)Conservative parallel transmit SAR limits were usedΦ matrix computed from forward and reflected power measures

RF Shim weights

Γ

- matrix

No Shimming

Φ

- matrix

Slide44

RF Shimming Weights Calculation

B

1+ map acquisition and field extraction: ~1 min

Φ matrix calibration: < 5 seconds

Maximum

efficiency weights calculation

: < 1

s

Slide45

Results: RF Shimming at 7 T

No RF Shimming

Maximum Efficiency RF Shimming

Transmit Efficiency (η)117286Measured Total Average Energy Deposition (Watts) 15558.8Mean Flip angle in ROI28.8° ± 10.7°25.1° ± 10.9°

Slide46

Any questions?

Slide47

See you after

Spring Break!

Slide48

Slide49

Slide50

Slide51

Slide52

Slide53


About DocSlides
DocSlides allows users to easily upload and share presentations, PDF documents, and images.Share your documents with the world , watch,share and upload any time you want. How can you benefit from using DocSlides? DocSlides consists documents from individuals and organizations on topics ranging from technology and business to travel, health, and education. Find and search for what interests you, and learn from people and more. You can also download DocSlides to read or reference later.
Youtube