Biomedical Engineering Reference
In-Depth Information
Subsequently, designs using more elements have been consid-
ered (for example, Daum and Hynynen 1999). Sasaki et al. (2003)
have described the use of a four-element array mounted in a
truncated spherical bowl that has been driven to produce a split
focus that gave a 3-4 times larger focal volume than obtained
with a single focus exposure.
A problem that arises with these regularly spaced phased-array
designs is the presence of grating lobes and other secondary max-
ima. Goss et al. (1996) showed that sparse, randomly distributed
arrays reduce the magnitude of these. Lu et al. (2005) described a
generic algorithm for optimization of phase and amplitude dis-
tributions for the reduction of grating lobes. There have been a
number of other studies of apodization and the use of subsets
of elements (Dupenloup et al. 1996; Gavrilov et al. 1997, 2000;
Hand and Gavrilov 2000, 2002; Filonenko et al. 2004). In a publi-
cation in 2009, Hand et al. (2009) described the performance of a
254 circular element, 1 MHz random array, capable of producing
five simultaneous foci. Melodelima et al. (2009) have described a
multielement toroidal transducer. This geometry allows the rapid
induction of large volumes of tissue within 7 cm of the trans-
ducer, and thus offers the possibility of rapid ablation of liver
tumors during liver surgery. In the search for a larger ablated
volume per unit time Chen et al. (2011) have proposed a cylin-
drical section single-element transducer design that is capable of
generating a line focus. This device is known as the “SonoKnife.”
The discussion to this point has focused on extracorporeal
transducers. There is also considerable interest in interstitial
heating sources. Here miniaturized sources are mounted on the
end of catheters or endoscopes. In this geometry, imaging and
therapy elements are placed at the end of the probe. The main
thrust of this has been in the development of ultrasound based
transrectal devices for thermal based therapies of benign and
malignant prostate disease (Diederich et al. 2004a,b; Lafon et al.
2004; Ross et al. 2004, 2005; Chopra et al. 2005; Kinsey et al.
2008). Miniaturized image/ablate probes for insertion directly
into the tumor of interest have also been proposed (Makin et al.
2005; Mast et al. 2011; Lafon et al. 2002, 2007).
5.2.5.3 phased arrays
In practice, focused beams are most commonly created using
phased arrays. These are usually formed with many indi-
vidual elements mounted on the surface of a spherical bowl.
Piezocomposite materials in particular lend themselves to this
type of transducer. Beam focusing and movement is achieved
by judicious choice of phasing of the drive signal to each ele-
ment. These have the advantage that the heating pattern can be
changed electronically, where appropriate, in response to feed-
back from monitoring of, for example, temperature or stiffness in
the tissue target. Electronic control of heating has the advantage
that the device/patient interface may be easier to provide, and
the maximum use of available acoustic windows can be made.
Where a single focal spot is desired, changes in phasing of the
element may allow an increased heated volume compared to that
of a similar single-element transducer. Phased arrays can also
be driven to provide multiple foci, which can then be scanned to
enlarge the heated volume. This approach is sometimes referred
to as multifocus scanning and was first proposed by Ebbini and
Cain (1989, 1991). The degree of flexibility in potential field pat-
terns is largely dependent on the number of elements available,
and their disposition on the surface of the transducer head. A
number of element geometries have been used, depending on the
application, with the largest arrays being used for HIFU treat-
ments of the brain (Sun and Hynynen 1998, Clement et al. 2000,
Pernot et al. 2005, Clement & Hynynen 2002, Pernot et al. 2003,
Hynynen et al. 2004, 2006, Aubry et al. 2007). An added advan-
tage of the phased array approach is the ability to use some of the
elements in dual imaging/therapy mode (Bouchoux et al. 2008).
Electronic adjustment of the phasing of each element of con-
centric annular arrays allows rapid movement of the focal region
along the beam axis (Huu and Hartmann 1982, Ibbini and Cain
1990). This is demonstrated in Figure 5.6.
The idea of using several closely spaced focal spots to obtain a
larger homogeneous temperature distribution by taking advan-
tage of thermal conductivity has been widely explored. Fan and
Hynynen (1995, 1996) modeled the fields from a spherically
curved transducer with a 4 × 4 array of square elements. They
showed that four focal points at a separation of 2-4 wavelengths
could be created. A maximum necrosed volume up to 16 times
that of a similar single-element transducer could be obtained.
5.3 acoustic Field propagation
During its passage through tissue, the energy contained in an
ultrasound beam reduces, that is, it is attenuated. The energy
reaching a given point in tissue is determined in part by how
much is scattered out of the main beam by tissue structures in
the beam path and by absorption. This energy loss is character-
ized for a specific tissue by its attenuation coefficient, which is
the sum of the absorption coefficient and the component due
to scattering as described below. Energy is also lost by specular
reflection and refraction of the beam. For a plane pressure wave
propagating in the x-direction, the pressure can be written as
5
0
-5
5
0
-5
5
0
-
-50
-40
-30 -20 -10 0
Axial distance from geometric focal peak (mm)
10
20
30
40
50
(
=
(
)
itxc
ω
−
/
Pxt
,
Pee
−
α
x
(5.6)
0
FIGURE 5.6
Field plots for an unsteered transducer (middle) and the
same transducer steered backward (top) and forward (bottom) along
the acoustic axis.
where α is the amplitude attenuation coefficient, ω is the angular
f requenc y,
P
0
is the pressure amplitude at the position
x
= 0, and
