BE280A 07 US2 - University of California, San...
Transcript of BE280A 07 US2 - University of California, San...
1
TT Liu, BE280A, UCSD, Fall 2007
Bioengineering 280APrinciples of Biomedical Imaging
Fall Quarter 2007Ultrasound Lecture 2
TT Liu, BE280A, UCSD, Fall 2007
Single-slit Diffraction
Source:wikipedia
!
Beamwidth "z
D
TT Liu, BE280A, UCSD, Fall 2007
Plane Wave Approximation
!
In general
U(x0,y0) =1
zexp j
2"z
#
$
% &
'
( ) F s(x,y)[ ]
k x=x0
#z,ky =
y0
#z,
Example
s(x,y) = rect(x /D)rect(y /D)
U(x0,y0) =1
zexp( jkz)D2 sinc Dkx( )sinc Dky( )
=1
zexp( jkz)D2 sinc D
x0
#z
$
% &
'
( ) sinc Dky
y0
#z
$
% &
'
( )
Zeros occur at x0 =n#z
D and y0 =
n#z
D
Beamwidth of the sinc function is #z
D
TT Liu, BE280A, UCSD, Fall 2007
Huygen’s Principle
Anderson and Trahey 2000
!
Wavenumber
k =2"
#
!
Oliquity Factor
r01
2
TT Liu, BE280A, UCSD, Fall 2007
Small-Angle (paraxial) Approximation
Anderson and Trahey 2000
r01
TT Liu, BE280A, UCSD, Fall 2007
Fresnel Approximation
Anderson and Trahey 2000
Approximates spherical wavefront with aparabolic phase profile
TT Liu, BE280A, UCSD, Fall 2007
Fresnel Approximation
!
U(x0,y0) =1
j"r01## exp jkr01( )s x1,y1( )dx1dy1
$1
j"z## exp jk z
2 + x1 % x0( )2
+ y1 % y0( )2( )( )s x1,y1( )dx1dy1
$exp( jkz)
j"z## exp
jk
2zx1 % x0( )
2+ y1 % y0( )
2( )&
' (
)
* + s x1,y1( )dx1dy1
=exp( jkz)
j"zs(x0,y0),,exp
jk
2zx02 + y0
2( )&
' (
)
* +
&
' (
)
* +
TT Liu, BE280A, UCSD, Fall 2007
Fresnel Zone
!
U(x0,y0) "exp( jkz)
j#z$$ exp
jk
2zx1 % x0( )
2+ y1 % y0( )
2( )&
' (
)
* + s x1,y1( )dx1dy1
=exp( jkz)
j#z$$ exp
jk
2zx12 + y1
2( ) + x02 + y0
2( ) % 2 x1x0 + y1y0( )( )&
' (
)
* + s x1,y1( )dx1dy1
=exp( jkz)
j#zexp
jk
2zx02 + y0
2( )&
' (
)
* + ,
exp( jkz)
j#z$$ exp
jk
2zx12 + y1
2( )&
' (
)
* + s x1,y1( )exp %
jk
zx1x0 + y1y0( )
&
' (
)
* + dx1dy1
=exp( jkz)
j#zexp
jk
2zx02 + y0
2( )&
' (
)
* + F exp
jk
2zx12 + y1
2( )&
' (
)
* + s x1,y1( )
-
. /
0
1 2
Phase across transducer face
3
TT Liu, BE280A, UCSD, Fall 2007
Fraunhofer Condition
!
k
2zx1
2 + y1
2( )Phase term due to position on transducer is
Far-field condition is
!
k
2zx1
2 + y1
2( ) <<1
z >>k
2x1
2 + y1
2( ) ="
#x1
2 + y1
2( )
For a square DxD transducer, x1
2 + y1
2 = D2 /2
z >>"D2
2#$D
2
#
TT Liu, BE280A, UCSD, Fall 2007
!
"z
D1
!
D1
!
"z
D2
!
D2
2
"
!
D1
2
"
!
D2
!
"z
Dz=D
2/"
= D
TT Liu, BE280A, UCSD, Fall 2007
Fresnel Zone
Anderson and Trahey 2000TT Liu, BE280A, UCSD, Fall 2007
c > c0
Focusing in Fresnel Zone
!
U(x0,y0) =exp( jkz)
j"zexp
jk
2zx02 + y0
2( )#
$ %
&
' ( F exp
jk
2zx12 + y1
2( )#
$ %
&
' ( s x1,y1( )
)
* +
,
- .
D
z0
!
expjk
2z0x12 + y1
2( )"
# $
%
& '
!
exp "jk
2z0x12 + y1
2( )#
$ %
&
' (
4
TT Liu, BE280A, UCSD, Fall 2007
Focusing in Fresnel Zone
!
U(x0,y0) =exp( jkz)
j"zexp
jk
2zx02 + y0
2( )#
$ %
&
' ( F exp
jk
2zx12 + y1
2( )#
$ %
&
' ( s x1,y1( )
)
* +
,
- .
Make
!
s x1,y1( ) = s0 x1,y1( )exp "jk
2z0x12 + y1
2( )#
$ %
&
' ( !
U(x0,y0) =exp( jkz0)
j"z0exp
jk
2z0x02 + y0
2( )#
$ %
&
' ( F s x1,y1( )[ ]
At the focal depth z =z0
Beamwidth at the focal depth is:
!
"z0
D
TT Liu, BE280A, UCSD, Fall 2007
Acoustic Lens
!
"z0
D= "F
D
c > c0
z0
!
U(x0,y0) =exp( jkz0)
j"z0exp
jk
2z0x02 + y0
2( )#
$ %
&
' ( F exp
jk
2z0x12 + y1
2( )#
$ %
&
' ( exp )
jk
2z0x12 + y1
2( )#
$ %
&
' ( s x1,y1( )
*
+ ,
-
. /
=exp( jkz0)
j"z0exp
jk
2z0x02 + y0
2( )#
$ %
&
' ( F s x1,y1( )[ ]
At the focal depth z =z0
TT Liu, BE280A, UCSD, Fall 2007 TT Liu, BE280A, UCSD, Fall 2007
Depth of Focus
!
When z " z0, the phase term is #$ = exp %jk
2z0
x1
2 + y1
2( )&
' (
)
* + exp %
jk
2zx1
2 + y1
2( )&
' (
)
* +
and the lens is not perfectly focused.
Consider variation in the x - direction.
#$ =kx
2
2
1
z%
1
z0
&
' (
)
* +
For transducer of size D, x
2
2=D
2
8
If we want #$ =,D2
4-
1
z%
1
z0
&
' (
)
* + <1 radian then
1
z%
1
z0
<4-
,D2.-
D2/
#z
z0
2<-
D2
The larger the D, the smaller the depth of focus.
5
TT Liu, BE280A, UCSD, Fall 2007
!
"z0
D= "F
D
z0
D
TT Liu, BE280A, UCSD, Fall 2007
Focusing with Phased Array
Anderson and Trahey 2000
TT Liu, BE280A, UCSD, Fall 2007
Focusing and Steering
Prince and Links 2005TT Liu, BE280A, UCSD, Fall 2007
Dynamic Focusing
Prince and Links 2005
6
TT Liu, BE280A, UCSD, Fall 2007
Doppler Effect
http://www.centrus.com.br/DiplomaFMF/SeriesFMF/doppler/capitulos-htmłchapter_01.htm
TT Liu, BE280A, UCSD, Fall 2007
Doppler Effect
!
"f =2vf
0
c # v$2vf
0
c
Examplev = 50 cm/sc = 1500 m/sf0 = 5 MHz
2vf0
c= 3333 Hz
TT Liu, BE280A, UCSD, Fall 2007
Doppler
http://www.centrus.com.br/DiplomaFMF/SeriesFMF/doppler/capitulos-htmłchapter_01.htm
TT Liu, BE280A, UCSD, Fall 2007
CW Doppler
http://www.centrus.com.br/DiplomaFMF/SeriesFMF/doppler/capitulos-htmłchapter_01.htm
7
TT Liu, BE280A, UCSD, Fall 2007
CW Doppler
f0 f0+Δf-f0-f0-Δf
*Effective width = 1/Tobs
TT Liu, BE280A, UCSD, Fall 2007
CW Doppler
!
Resolution "f =1/Tobs
"v =c"f
2 f0
=c
2Tobs f0
ExampleDesign goal : "v = 5 cm /s; f
0= 5 MHz
Tobs =c
2"vf0
=1500m /s
2 0.05m /s( ) 5 #106( )= 3ms
Note that for a depth of 15 cm, it takes only200 usec for echos to return.
TT Liu, BE280A, UCSD, Fall 2007
PW Doppler
http://www.centrus.com.br/DiplomaFMF/SeriesFMF/doppler/capitulos-htmłchapter_01.htm
TT Liu, BE280A, UCSD, Fall 2007
PW Doppler
http://www.centrus.com.br/DiplomaFMF/SeriesFMF/doppler/capitulos-htmłchapter_01.htm
8
TT Liu, BE280A, UCSD, Fall 2007 Siemens Medical Systems; jnormal common carotid artery TT Liu, BE280A, UCSD, Fall 2007
Aliasing
!
Measure Doppler shifts at a specified range
For unambiguous range, one pulse at a time.
TPR =2rmax
c (e.g. 200 usec for 15 cm depth)
To avoid aliasing require
1
TPR
> 2"fmax
Δf-Δf*
1/TPR
=Δf-Δf
AliasedΔf-Δf
TT Liu, BE280A, UCSD, Fall 2007
Aliasing
http://www.centrus.com.br/DiplomaFMF/SeriesFMF/doppler/capitulos-htmłchapter_01.htm
TT Liu, BE280A, UCSD, Fall 2007
Aliasing
http://www.centrus.com.br/DiplomaFMF/SeriesFMF/doppler/capitulos-htmłchapter_01.htm
9
TT Liu, BE280A, UCSD, Fall 2007
Aliasing
http://www.centrus.com.br/DiplomaFMF/SeriesFMF/doppler/capitulos-htmłchapter_01.htm
TT Liu, BE280A, UCSD, Fall 2007
PW Doppler
!
Velocity Resolution (same as with CW)
Tobs >1
"f=
c
2"vf0
Range Resolution
Want to interrogate velocities from a small region "z =cTpulse
2
We also need to make sure that particles remain within this region
over the observation time Tobs
vmaxTobs < "z# Tobs <"z
vmax
=cTpulse
2vmax
TT Liu, BE280A, UCSD, Fall 2007
PW Doppler
!
Design Example
Rmax = 6 cm " TPR =2(0.06m)
1500m /s= 80 µsec
1
TPR
> 2#fmax =4vmax f0
c
c
4TPR f0
> vmax " for f0 = 5MHz we find that vmax < 93.75cm /s
If we choose #v =1cm /s then Tobs =c
2#vmax f0
=15ms
Range resolution : #z > vmaxTobs =1.4cm
Tpulse =2#z
c=18.8usec