Transcript MRI Perfusion-Sensitive Imaging - Dartmouth

Advances in Imaging:
Echo, CT, CMR
Justin D Pearlman MD ME PhD
Director, Dartmouth Advanced Imaging
Center
Disclosures
• Consultant for:
–General Electric
–Picker/Marconi/Phillips
–Chiron
–Boehringer-Ingelheim
–MagnaLab
–Perfusion=off-label use of
contrast
Dartmouth Advanced Imaging
Center - Aims
High-end imaging capabilities; Bench->Bedside
•1. Realtime CMR •Viability
•Dx
•2. 4D Cardiac CT
•Myopathy
•Rx
•3. 3D Echo, PET
•Microcirculation
Equipment
Echo
MR
CT
Images: CMR vs. Echo
CMR short axis
Cost: $500-$1500
Echo short axis
Cost: $400-800
Similarities
Echo
• Inject packets of
energy waves,
pulsed,
Receive echoes
• Scan to collect data
to convert to image
• Tomographic
• Dynamic
• Flow signal from
CMR
• Insert packets of
energy waves, pulsed,
Receive echoes
• Scan to collect data to
convert to image
• Tomographic
• Dynamic
• Flow signal from phase
shift
Differences
•
•
•
•
•
•
Echo
Sound Speed
– 1540 m/s
Echo=A-mode (amp-time)
Stopped by
– Metal
– Bone
– Air-tissue
Views limited by rib window,
contact, angle
Resolution depends on
frequency, beamwidth
Bright blood requires
contrast
•
•
•
•
•
•
CMR
Radiowave Speed
– 299,792,258 m/s
Echo=K-mode (amp-spatial
freq)
Distorted by
– Metal
– (No problems with bone,
air/tissue)
Any view
Resolution is adjustable down
to 10 microns, limited by noise
and acquisition time
Bright blood many ways
Basis for MRI
B0 + Gx,y,z
+
B1
+
Mz  Mxy
Magnetism
Magnetism
Longitudinal
Magnetization
Change in Magnetization
1
0.8
T1/TR
1-e(-TR/T1)
0.6
0.4
0.2
0
-1
0
1
2
3
4
5
Time (seconds)
Longitudinal
Magnetization
1
1-2e(-TI/T1)
0.5
T1/TI
0
-1
0
1
2
3
-0.5
4
5
Time (seconds)
Transverse
Magnetization
-1
1
T2/TE
e(-TE/T2)
0.5
0
-5
5
15
25
Time (milliseconds)
35
45
K-Space
K-space sums
3,2
(3,2) + (2,5)
+
2,5
(3,2) + 0.5 (2,5)
=
Sum
0.5 (3,2) + (2,5)
Fourier Transform
Pulse Sequence
Gradient Echo (FLASH)
Pulse Sequence
Inversion Recovery (STIR)
Methods
• Magnetization Preparation
• Excitation
• Spatial Encoding, Echoes
• Image Reconstruction
Bright Blood TurboGradient Echo, Dark Blood FSEDIR, Fat-suppresive TIR, …
(100’s)
Corresponding notions
• Echo Intense
• Short T1 (if T1weighted image)
• Short T2 (if T2weighted image)
• Water (if fat
suppressed)
• Fat (if not fat
suppressed)
Fat vs. Fat Suppression: RVD
Fat+
RV
vs.
Fat-
RV
JDP 2/02
SMART function
40%
40%
SMART
bFGF-2
Fixed
30%
30%
Saline
20%
20%
10%
10%
0%
0%
Motion
Motion
Thickening
Thickening
Pearlman JD et al Serial motion assessment by reference tracking (SMART): application to detection of
local functional impact of chronic myocardial ischemia. J Comput Assist Tomogr, 2001. 25(4): p. 558-62
Self-Triggered MRA
Coronary Imaging
Coronary Imaging
Calcium Scoring
Claims:
•Negative score may
indicate non-cardiac
sources of chest pain
•Scores over 1,000 predict
coronary event within the
next 2-3 years
•Positive scores referred
for catheterization or
stress test
•BUT significant disease
may have negative score
•Positive score may be
stable plaque
Current Use:
Aberrant Coronary Origins
Dynamic CT
Elastic Match of Coronaries
• Fast CT of mom
• Elastic match 
contrast
• Simulated
holography as
background, for
context
Coronary Sinus Rx
4D CMR
Perfusion-Sensitive Imaging
Resting delayed blood arrival predicts ischemia
Rest Delayed Blood Arrival
Dark
Late Zone
Arrived
Space-Time Map
We introduced SpaceTime Maps to see delay
in blood arrival in a
single derived image
Perfusion Equitime
Rest MRI vs.
rest Thallium / stress MIBI
Table 1: Clinical Characteristics of Study Population
Patients
N
Disease vessel
105
%
38
P=0.35
Prevalence
97
2.69
0.10
CABG
42
10
26
1.08
0.13
Angioplasty
36
22
56
0.92
0.14
Stent
15
11
28
0.38
2.10
2
1
0
MRI
0.11
20
15
10
5
20
17
12
10
12
10
4
99
100
17
9
8
3
25
23
rest
P=0.01
30
stress
25
Rest
P<0.001
P=0.43
3 3 4
5
2 2 2
Coverage (%)
Number of Defects
MRI
30
Stress
Coverage of Defects
33 33
32
3.13
2.97
3
Agreement between Rest MRI and Rest / Stress Nuclear
35
P<0.001
4
SE
# Segments (of 8)
Total number
Blood Distribution Defects
80
75
83
81
61
62
str/mri res/mri mri/res
res/str
60
40
20
0
0
1
2
3
4
5
Location
6
7
8
mri/str
Target/Reference
str/res
Viability
Delayed Enhancement
62 year old patient with 3-vessel CAD c/o angina at rest.
Hx MI 1992, PTCA LAD 1992, CABG 1995.
Scintigraphy, MRI : lateral + anteroseptal wall defects
Scar
RV
TV
septum
RA
LVOT
LV
lateral
LA
MV
Delayed Enhancement vs.
Delayed Arrival
Molecular Imaging
bFGF2
Microvascular MRI
• Tissue
bright
• Major vessels visible
• Dynamic physiology
• Small
vessels hidden
Angiogenesis-Sensitive MRI
r=.95
No contrast
Nat Med 1:1085 ‘95
Dark Flash
3D CT Validation
Radiology 214:801 ‘00
Acad Rad 4:680 ’97
Dark Flare Predicts Improved Blood
Arrival From Angiogenesis
Baseline
1 Month
2 Months
Dark Flare/Delayed Arrival Combined
First Dose-Response for Angiogenesis Rx
DA=Demand, CX=Response
Angiogenesis imaging may also
help diagnose and treat cancer
34 y.o. woman with a
palpable breast mass.
Ultrasound negative
Mammography negative
Collateral Sensitive MRI:
fat black, collateral
neovascular development
flashes; cancer found.
MRI Microscopy in Large Target
Limit signal to 1 cm2
Fold-over problem
Avoid fold-over
Look at bowl of kiwi
40 micron resolution
Fold-over
RME =
ResponseModulated
Excitation
No fold-over
Intravascular Imaging
What to know
• Vocabulary
–B0, B1, Mz, Mxy, T1, T2, T2*,, ,  , 
–TI, TR, TE, , Matrix, FOV, , TD, TW
–GE, SE, FISP, HASTE, …
• Tilted Tomographic Anatomy
• Pathophysiology, Clinical Decisions
• Physics, Image Processing
Echo vs. CMR
CMR
Echo
“Both are watching out for the CAT skinner”
Clinical Example
 42 y.o. man with
large cell
lymphoma
 Radiation to chest
 Paroxysmal atrial
fibrillation
 CT: Mediastinal
mass ? LA
compression
Long Axis 4 Chamber View
Echo
MRI: mass, effusion
Long Axis 2 Chamber View
Echo: ? NL fxn
MRI:effusion,mass
Long Axis 3 Chamber View
Echo
MRI
Long Axis 5 Chamber View
Echo: WNL
MRI: Effusion,Mass
Short Axis Cine (Stack) View
Echo: WNL
MRI: Effusion, Mass
Short Axis Stack Cine
Clinical Example
• 63 y.o.
Woman
• Paroxysmal
Atrial Fib
Long Axis 2 Chamber View
Echo: LVH
MRI: LVH
Long Axis 3 Chamber View
Echo: ASH
MRI:LVH+RVH
Long Axis 4 Chamber View
Echo: ASH; lung/RV
MRI: LVH+RVH
Long Axis 5 Chamber View
Echo: “ASH”
MRI: LVH+RVH
Cine Stacks: RPA stenosis
Short Grid Base Cine View
Echo: hyperkinetic
MRI: Rt septum hypo, order
Clinical Example
• 61 y.o.
woman
• Tamoxiphen
Rx
• Idiopathic
CHF
Long Axis 2 Chamber View
Echo
MRI
Long Axis 4 Chamber View
Echo
MRI
Long Axis 5 Chamber View
Echo: ?Good EF
MRI: EF 11%
Long Axis Rotating Views
MRI: DCM, CS,  IVC
Short Axis ( Grid Tag Stack) View
Echo: Low EF ?Constriction
MRI:DCM, No constriction
Clinical Example
• 60 y.o. Woman
s/p L
Mastectomy, XRT
• CO=1.0
• TR=4+
4L Chamber View
Echo
MRI
Grid Cine vs. Dobutamine
0 ug/kg
10ug/kg
20 ug/kg
RV Strain vs Dobutamine Rx
14.00
13.00
0 ug/kg
10 ug/kg
20 ug/kg
12.00
Strain
11.00
10.00
9.00
8.00
7.00
6.00
0.09
0.17
0.26
0.34
0.43
0.52
Seconds
0.60
0.69
0.77
0.10
Apical RV Strain vs. Rx
14.00
13.00
0 ug/kg
10 ug/kg
20 ug/kg
Strain
12.00
11.00
10.00
9.00
8.00
7.00
6.00
0.09 0.17 0.26 0.34 0.43 0.52 0.60 0.69 0.77 0.10
Seconds
AV Prosthesis
Suture Dehiscence
Aortic Cusp Aneurysm
MR Artifacts
• Chemical Shift Artifacts : Fat, water yield sum of two shifted images 3.5 PPM
• Aliasing : Field of view divides all of space – sine wave is infinite pattern, so copies
sum
• Black Boundary Artifacts : At 1.5 T, 3.5 PPM water - fat shift cancels at 4.5 ms
multiples from 2.3 ms, eg 6.8, 11.3, and 15.9 ms. Avoid with TE's close to 4.5, 9,
13.6,....
•
•
•
•
•
•
•
•
•
•
Gibbs or Truncation Artifacts : Ringing
Zipper Artifacts : Door open
Phase-encoded Motion Artifacts : Ghosts
Entry Slice Phenomenon : Artery/Vein dark or bright by slice order; false
“clot”
Slice-overlap Artifacts : Faded
Magic Angle Effects : Tendon gets T2 increased 100x at 55 angulation
Moire Fringes : Aliasing + Phase differences R/L
RF Overflow Artifacts : Washed out
Central Point Artifact : Spike
Susceptibility Artifacts : Microscopic gradients -> Bright/Dark spots
Thank you for your
attention
!
Key Terms
•
•
•
•
Resonance = Specific matching frequency
Excitation = Sending in a pulsed radiowave
K-space trajectory = Data collection pattern
Magnetic Field Gradients =
– Spatial encoding tool
• MR Echo is not from tissue interface – it is
externally produced from all locations in slice by
– Refocusing Radiowave Pulse
– and/or Gradient Reversal
• Image is produced by “Fourier Transform”
– Converts “K-space” data to “X-space” image
• Pulse Sequence = Sequence of pulses, gradients
and steps to get an image or series of images
Key Terms
• T1 = Time to develop magnetization
• Gradient = Magnetic Field Slope low-high in X, Y, or Z
• Gradient-echo = echo caused by reversing X
gradient
• Spin-echo = echo caused by addition of a radiowave
refocusing pulse
• T2 = Time constant for loss in transverse
magnetization with spin-echo
• T2*= Time constant for loss in transverse
magnetization with gradient-echo; susceptibility
• TR = “Repetition time” = T1 contrast weight
• TE = “Echo time” = T2 or T2* contrast weight
• T1 Weighted = method emphasizing T1 differences