Biomedical Engineering Reference
In-Depth Information
rough skull with
phase correction
In water
rough skull
5
5
5
0
0
0
5
5
5
-5
0
5
-5
0
5
-5
0
5
5
5
5
0
0
0
5
5
5
-5
0
5
-5
0
5
-5
0
5
x (mm)
FIGURE 14.2
(Top row) Lateral pressure profile from a hemispherical transducer (
f
= 306 kHz) in water (left column), in water though a human
skullcap (center column), and through a human skullcap with hydrophone-based phase correction (right column). (Bottom row) Lateral pressure pro-
file from a hemispherical transducer (
f
= 830 kHz) in water (left column), in water though a human skullcap (center column), and through a human
skullcap with hydrophone-based phase correction (right column). (Based on data from Song, J. and K. Hynynen,
IEEE Trans Biomed Eng
57, 2010.)
focus, but a higher level of control can be achieved using a spa-
tiotemporal inverse filter that allows constraints to be placed on
the acoustic field around a focal point providing superior focus-
ing capabilities [Tanter et al., 2001]. Placement of a hydrophone
at the focus is not practical for
in vivo
work, and while time-
reversal could be achieved by placing an ultrasound emitter in
the brain during biopsy, it is still an invasive technique.
Hynynen and Sun [1999] proposed an entirely noninvasive
method for correcting skull-induced distortions of the acous-
tic field. They proposed to use MRI information to model the
shape of the skull bone and calculate the required element delays
by simulating sound propagation from a virtual source. Their
model consisted of three homogenous layers: water, bone, and
brain tissue. This model was expanded to assign homogeneous
material properties based on CT density information [Clement
and Hynynen, 2002]. Aubry et al. [2003] further advanced this
technique by using CT derived density information to produce
a model with heterogeneous material properties. Simulation
based focusing can be achieved using pre-procedure CT image
information and registering it with MR images of the patient
during the procedure.
Although simplified models can be used to perform focusing
in real time for clinical treatments, more precise simulation-
based focusing can be time consuming. To provide a quick and
more precise, noninvasive focusing method, a further extension
of time-reversal focusing was proposed where cavitation bub-
bles act as the acoustic source [Kripfgans et al., 2000; Pernot
et al., 2006]. Acoustic droplet vaporization could be used to
generate bubbles at the focus [Kripfgans et al., 2000], or the
cavitation bubbles can be induced at the focus in the absence
of droplets using a very short pulse [Pernot et al., 2006]. The
recorded acoustic signature is used to calculate time delays in
the same manner as if a transducer were placed at the focus. The
phase corrections are then applied to maximize focal efficiency.
Recent work using this technique restored 97% of the focal pres-
sures that can be achieved using hydrophone-based correction
through a single
ex vivo
human skull in water, compared with
83% using only CT-based correction methods [Gateau et al.,
2010]. This technique has produced good results, greatly reduc-
ing focusing time over full-scale simulation-based techniques,
but its safety has yet to be examined
in vivo
where cavitation
thresholds are highly variable.
In addition to phase correction techniques, White et al.
[2005] examined two techniques for amplitude correction,
which, combined with phase correction, can be used to com-
pensate for losses through skull and achieve better focal recon-
struction, but with the possibility of localized skull heating.
Additionally, it was shown that a multifrequency approach can
result in an increase in focal intensity over brief periods [White
et al., 2006].
