12 PHYSICS OF MAGNETIC RESONANCE 0

CHAPTER

resonance NMR
Nuclear magnetic
atoms
Teoncerns the nuclei of
radioactivity
not related to

microscopic magnetization or neutrons

have

Atoms with oddmomentum of

number protons
Io
an angular called spins moments
atoms are
these
in MRI

qtcmswamomatumqqqyym.se

F r
Tgyromagnetic ratio Crps tesla

Hz Hesla
F IT

FERB
O.OOl 909
t
b B field applied aligned
No field
bulk magnetization
Bulk magnetization with B
No

Ns
TU E fun
Tmagnetic moment of
bulk magnetization each spin
vector
 a
apply Bo Bo 2
If we

Mo Bo 8242 Pp
4kt

T gyromagnetic ratio
constant
h Planck's constant
Boltzmann's volume
k
of spins per unit
proton
density number
PD

on the type of spin nucleus

T depends
7 MHz Tesla
Nucleus MRI mostly
i 42 58 images hydrogen
H abundant in Hz
10.71
13C
19
f
40.05
3 11 26
p
NT N'crit
time and position
NT is a function of
Also def ne I Ion
momentum
B Uk angular longular momentum
of each spin

So ri r 5
VE
f macroscopic
microscopic
an ensemble of spins
 and Larmor Frequency
Precession

DITCH RICH B Ct
dt
Tu 85
using
Ct
did T NTH x
dt

B Ct Bo I
if Mo sin x cos O
2 M o
Mo sin x sin O
My o

Mzco Mo cos
f Bo I

Mo Sina cosC TBot 10

Mx Ct
Mo sinx sin C TBot 10
My Ct
No cos
Mz t
No INTlo
Where
Wo in
about Bo with frequency on
Precession
TBo Larmor frequency in rps
Wo
in Hz
f Bo Larmor freq
Vo the resonancefrequency
also called

so No sinx cos C 27kt t0 clogottwainfeon

µ c Sin 2M Not 0
Mo Sina
MyCt
t No cos X
Nz
Vo 42.58 MH 1.5 Tesla
MRI Scanner Tesla
1.5 T
Bo 1.5T 63.87 MHZ
3T
MII
x
scanner yo 42.58
3T MRI
127.74 MHz
In an MRI scanner Bo must be VERE homogeneous

3T scanner
max 1 ppm variation e.g for a
variation in field
parts permillion
Max 3Mt
a field
to generate such
VERY difficult are expensive
MRI scanners
reason why

magnetic susceptibility

t.i.it
H
HiFii
il ill 17111
B lines
B lines dense are denser
are less

Chemical shift
Bo _Bo CI g
to Vo h G
constant in ppm
shielding
down 3.35ppm
CCHz is shifted by
Hydrogen
in fat in water H2O
with rescept to Hydrogen
3 35 10
DV Vo 63.87MHz x
atl.SI Hz difference
214
6
3.35 10
DV Vo 127.74 MHz x

at3T 428 Hz difference

Transverse and Longitudinal magnetization
at along
Longitudinal magnetization static
fB
as
A the same direction

JI l 9 Bo field i e Ma Ct
I in the
Transverse magnetization
the
Y orthogonal to
plane i e MxyCt
dM y static field
x

Mx CH j MyCt
Mxyct or phaseangle
tan My_ phase
0 Nx

Mo Sina cos 81301 0

Previously Mx Ct
Mo Sina Sin 8Bot 0
My Ct
No cos
Mz t
JKBot 0
Mo sing e
Now Mxyct J 21TVo t 0
Mo sing e

magnetization
rotates rapidly coil MRI
Transverse pickup
can be measured with a
signal
This
it

pickup coilto charges in Mx

sensitive
d
EMF skip steps

2tivot 0
V t IVI Sin
in
MRI signal
magnitude of
 Where IVI Zit Vo Vs Mo Sina Brc coil sensitivity
Larmor
Tvolume
ofthesompleT magnitude of
frequency or voxel size magnetization
in Xy

M y
signal is proportional to
so in MRI
BOZ
d Bo Vo X Bo
So signal
Mo

Signal is maximum
when X
IL
where is called tip angle or flip angle

4 this
we control
in a reference
Rotating
frame express magnetization
that rotates at
coordinate system
the Larmor frequency
2TVot
X x cos 2Tivot y sin
2tVot
X sin 27kt y cos
y
z 2

Then
Ct Mo sina et
May
Tappears stationary in rotating frame
RF Excitation
in pickup co
M induces signal
Transverse component of M
creates transverse componentof
current through a coil
T applied at the resonance
Larmor frequency
c

t laboratory frame
in
B cosC 2tivot I
B sin C 2HVot y
BT
R
ff rotating fame
in

ftp.ez.T BT Bi I
TTU x B I
dj
it
i
f ffI H D
c
c

B Ct

t i I
c
i
14 flip angle
So
713,212
envelope is time varying
In general if
Bicton Belt

I I B.ec sinC 2ttVot y

B Ct B ect cos C 2tivot

BT t B ect I
 Then
1
x B e Ct DE flip angle
To a.k.a tip angle

excitation pulse
L
Iz
is typical
Mxy
Induces maximum transverse component
maximum signal

want to use a 3ms excitation pulse

Example We
What should be the amplitude
with a If
Rf excitation
of 3ms
21T F B Tp 2T 42.58MTHI Bi
X Iz
6
Tesla l 96MT 0.0196 Gauss
B l 96 10

10
4 Tesla O 1 MT very small Rffreld
1 Gauss is sufficient to
OR 1 MT 10 Gauss magnetizat
tip the

Relaxation
Relaxation a.k.a spin spin relaxation
1 Transverse causes the received signal to decay
the magnetic
due to perturbations in
nearby
field from other spins
coherence
loss of it
na
n't
II tip att
it
0

Fi Tad y
f y
time passes Random dephasing
time passes
loss of
due to field coherence
perturbations
JG'T t 0 Hz
Mo sin e e
MxyCt
t T called Ts decay
Mx'y Ct Mo sina.ee e or
12 relaxation
 signal R envelope

s
t

t induction decay FID

free
i
i
t
Tz decay

Tz decay perturbations
due to further outside freld
relaxation
Relaxation Ca k a spin lattice
2 Longitudinal back to its
recovers
MzCt Mo exponentially
equilibrium value
Rf pulse with
Mo Cosa after
suppose MzCt Ot flip angle x

Tha th
th
Mo l e Mz ot e
MaCt
Maltin
Mo
Mzcoy
t

different tissue types

Ti T2 are different for
creates main contrast in MRI images
the
all materials
and T Z Tz for
C T 2500ms
Typically 250 ms
25 ms L Tz L 250 ms

in equilibrium if no RF pulse
So a sample is
at least 3T Max
has been applied for
 is in equilibrium and a
Iz pulse
Exampley suppose a sample
What happens to longidutinal
applied at t
O
is
magnetitation
O
Mz Ot Mo cosit 2
th
Mo l e
501 Mz Ct

How about transverse component

Mx Ot D
Sin Tiz Mo
My Ot Mo
tha
So My t Mo e
Mo.eJTk tha
e
Alternatively Maj t

The Bloch Equations

and relaxation
both the forced
Putting together what is called Bloch equations
behavior gives
RICH MT
ditch ZNTCH Bct x R
Tt
TU in laboratory frame
BT RT Ct
t

R T2 Relaxation matrix
y
y
Spin Echo

i
x
i
1 aiongPY.se

Z at
n't

E Timepasses
t
es

Pulse Sequence Diagram

adf.h.EE
T FET
TE echo time

outside perturbations
due to
spin echo recovers dephasing relaxation effects
but cannot recovers inherent Tz
 E e Two 1H spins are in different locations
and there is 2ppm
in a 1.5T MRI scanner
difference between the two locations
field be 1800
How long will it
take these spins to
out of phase 6
I 5 2 10
Dr FD Bo 42.58 MH
127.7 Hz
b a 3.9ms
IT Dt
zit DV At DV
q time
relatively short

their difference at TE 2 if
what will be phase
TE 40 ms
3
a 16 radians
21T x 127.7 x 40 10
21T DU TEI
coherence
they completely lost

contrast mechanisms
have T 12 Pip
tissues different
Different proton
density

simple mR
sequence
in
control tip angle
control TE echo time
time
control TR repetition 180
800
wait
µi N
RF i
W 1 S's
l
I 1 gthsifgnisanthat
MR i 1 we w 4
signal fl 1
ss
K TE

TR
 At 1.5 T
Relative Pp Ts Cms Ti ms
matter 0.61 67 510
white 760
matter 0.69 77
Gray 280 2650
Cerebrospinal fluid CSF 1.00

Contrast TE Long TR
Weighted short
Proton Density
1H spins in the sample
Goal image intensity
X
of or voxel

maximumsignal
2 412 for in equilibrium
n 3T max

wait till everything is 15 20ms

TE
i TR 26000ms 4
Typical almost all thespins not much to
back to
relaxed
fullylongitudinal direction decay when
echo is acquired
Rf pulse
before the next

TR Mo
µz Co Mz TE T
J2HtoTE.e
TE Mo e.JO e
Mxy to T2
TE is small compared
Mxy TE
No if
Mo proportional to PD
MR signal
x IN xy TE
TE
TR medium length
Contrast Long to Tz
Tz weighted comparable

TR 6000 ms
e g
TE looms
a T 2

Mz TE Mo
Mz o TEH
JO 52mV TE e
Mxy TE Mo e e
T2
x 1M y TE Mo e
MR signal
 medium length TR short TE
T weighted Contrast T
comparable to
e
HT
Mz Ot e HT
Mz Ct Mo l
Remember Assume the pulse
1 0
then Mz 0
Mo l e th

TR X T
wait for TRH
Mo l e
Mz TR
IEP'fuse tri
µz THT JO
Mo l e
Mxy TRT TRH eJ0eJ2tVoTE
TEH
Mo l e e
Mxy TR 1 TE
L
if TE Cc Te

TRH
IMxy TR c TE Mo l e
MR signal T decay
weighted by
we will get the
same Mxy at
TE for N 10
NTR

e g TR 2600ms
TE 215 20ms
a 1712