Webb Andrew Technical challenges and clinical research

Report
Technical challenges and clinical research applications of
ultrahigh field MRI
A.G.Webb
Professor, Director C.J.Gorter Center for High Field MRI
Department of Radiology
Leiden University Medical Center
Leiden, The Netherlands
Outline
Why a very high field scanner? What can it do?
Why doesn’t it work and give nice images automatically?
How do we address the major challenges?
Why do we need the input of medical physicists?
Clinical applications and future input into radiotherapy
2
Philosophy
Σα βγεις στον πηγαιμό για την Ιθάκη,
να εύχεσαι νάναι μακρύς ο δρόμος,
γεμάτος περιπέτειες, γεμάτος γνώσεις.
3
7T system
~50 worldwide
Why ultra-high field MRI?
Image quality is proportional to magnetic field strength
Signal to noise at 7 tesla 2.3 times higher than 3 tesla
Higher resolution and faster (for patients) MRI
Improved sensitivity to diffuse iron deposition (neurodegeneration)
Intrinsically better angiography to visualize small vessels
Increased spectral resolution for metabolic studies
Congratulations on purchasing your new Philips 7T
- a bargain at €8,500,000
Compared to your old 3 Tesla Philips is delighted to offer significant
increases in……..
Image non-uniformities
Potential for heating the patient
Questions about safety/implants/dental wires
Motion sensitivity
Difficulties in image segmentation
Complexity of cardiac triggering
But also significant decreases in
Number of RF coils commercially available
6
The first technical challenge –
design of customized detectors
Image non-uniformities
Potential for heating the patient
Questions about safety/implants/dental wires
Motion sensitivity
Difficulties in image segmentation
Complexity of cardiac triggering
8
Image non-uniformities at high field
3T
4T
5T
6T
7T
9
8T
10T
12T
General observations at high fields
Dielectric constant
• Overall, RF wavelength in tissue decreases with B0 field strength
 tissue 
1
f
r
Relative dielectric constant
Wavelength (cm)
Muscle
60
75
50
70
65
40
60
30
55
20
50
10
45
frequency (MHz)
frequency (MHz)
10
RF inhomogeneity
constructive/destructive interference
~12 cm
11
General observations
high fields
Solution 1 - multiple
transmitatchannels
RF waveform
generator 1
Power
amplifier
Tx/Rx switch 1
Coil element 1
Digital receiver 1
RF waveform
generator 2
Power
amplifier
Tx/Rx switch 2
Coil element 2
Digital receiver 2
RF waveform
generator 3
Power
amplifier
Tx/Rx switch 3
Coil element 3
Digital receiver 3
RF waveform
generator N
Power
amplifier
Tx/Rx switch N
Coil element N
Digital receiver N
12
The alternative and slightly cheaper method
New, high permittivity materials
How do dielectric materials work?
Displacement currents in
the dielectric material
produce a secondary
local RF field which
increases the total B1+
14
Dielectric pads in imaging
FLAIR
normal
with pads
TSE
Abdominal imaging at 3 Tesla
(a)
(b)
(c)
(d)
(e)
(f)
Image non-uniformities
Potential for heating the patient
Questions about safety/implants
Motion sensitivity
Difficulties in image segmentation
Complexity of cardiac triggering
17
ConductivityGeneral observations at high fields
Conductivity increases with frequency
Conductivity of gray matter (S/m)
0.55
 eff   io n   Im   r  0 
0.50
0.45
P=1/2 E2
0.40
0.35
100
150
200
250
300
350
frequency (MHz)
18
400
RF inhomogeneity
constructive/destructive interference
~12 cm
19
Increased SAR and heating at 7T
3T 128 MHz
128 MHz
7T 300 MHz
300 MHz
Temperature rise (oC)
SAR (W/kg)
20
How do you General
ensureobservations
safety? at high fields
The RF engineer is the first person
to be tested!
21
Call upon the
medical
physicsatspecialists
General
observations
high fields
Requires flexibility
22
General observations at high fields
lack of self-awareness
23
General observations at high fields
Attention to detail
24
General observations at high fields
Rigorous safety testing procedures
25
Electromagnetic simulations
Phantom heating tests
SARpoint (W/kg)
10
1
0.1
Image non-uniformities
Potential for heating the patient
Questions about safety/implants/dental wires
Motion sensitivity
Difficulties in image segmentation
Complexity of cardiac triggering
27
Reduction in image quality in patients
High quality obtained in volunteers is typically not reproduced
in AD patients
Healthy volunteer
AD patient
28
In vivo results of f0 fluctuations
freq (Hz) / A.U.
20
resp. belt
nav. phase
10
0
-10
50
0
Normal volunteer
100
50
time (s)
150
150
before correction
frequency (Hz)
20
10
0
-10
250
0
AD patient
300
50
time (s)
350
150
29
On-line monitoring of frequency variations
Some examples
Image quality is significantly
improved
31
Image non-uniformities
Potential for heating the patient
Questions about safety/implants/dental wires
Motion sensitivity
Difficulties in image segmentation
Complexity of cardiac triggering
32
Reduced contrast makes segmentation difficult
T2*-magnitude
T2*- phase
T1
Specialized image segmentation algorithm
Image non-uniformities
Potential for heating the patient
Questions about safety/implants/dental wires
Motion sensitivity
Difficulties in image segmentation
Complexity of cardiac triggering
35
Problems with cardiac triggering
Overwhelming magnetohydrodynamic effect
Develop acoustic triggering
Principle developed by Niendorf group on Siemens 7T platform
Commercially available for mildly ridiculous price
Develop an open-source Arduino-based system for continuous
Improvement amongst users
Develop acoustic triggering
Technical “solutions”
High permittivity materials
Accurate SAR modelling
On-line “motion” monitoring
Acoustic cardiac triggering
Phase/magnitude image segmentation
7T Cardiovascular MR
Coronary MRA
7T Cardiovascular MR
Coronary MRA, 7T versus 3T
S.G.C.van Elderen, M.J.Versluis, J.J.M.Westenberg, H.Agarwal, N.B.Smith, M.Stuber, A.de Roos and A.G.Webb,
van Elderen et al., Radiology 2010;257:254-259
In vivo coronary magnetic resonance angiography at 7 Tesla: a direct quantitative comparison with 3 Tesla,
Radiology, 257, 254-259, 2010.
7T Cardiovascular MR
Ischemic Cardiomyopathy, RCA
General
observations
at high fields
Carotid artery
vessel
wall imaging
T1
T2
TOF
Cochlear imaging
MIP
Inner ear imaging – cochlear implants
3T
7T
Musculoskeletal applications of 7 Tesla
High resolution imaging of the human vertebra
Ankylosing Spondylitis
•Inflammation in spine and sacroiliac joints
Water/fat images of sacroiliac (SI) joint
High resolution imaging of the eye
High resolution imaging of the eye
Uveal melanoma patients
ultrasound
Proton beam therapy planning
Acknowledgements
Itamar Ronen
Hermien Kan
Maarten Versluis
Thijs van Osch
Sanneke van Rooden
Ece Arcan
Francesca Branzoli
Sebastian Aussenhofer
Eidrees Ghariq
Wouter Teeuwisse
Mark van Buchem
Wyger Brink
Paul de Heer
Jeroen van der Grond
Aλλά μη βιάζεις το ταξείδι διόλου.
Καλλίτερα χρόνια πολλά να διαρκέσει·
και γέρος πια ν’ αράξεις στο νησί,
πλούσιος με όσα κέρδισες στον δρόμο,
μη προσδοκώντας πλούτη να σε δώσει η Ιθάκη.

similar documents