Biomedical Engineering Reference
In-Depth Information
attenuation of up to 2.4 dB/cm at 3.5 MHz were found in porcine
liver after heating to 80°C in 300 sec.
109
Ribault and coworkers also looked at the effect of temperature
rise on frequency-dependent attenuation and found that tissue
damage (lesion formation) caused a change in attenuation in
porcine liver
in vitro.
5
This effect was found by looking at the
backscattered signal over the volume of the lesion and compar-
ing the power received before and after high intensity focused
ultrasound (HIFU). Other investigators have observed similar
effects in the past.
111
Thus attenuation is of interest for thermom-
etry and may have application with appropriate processing at
temperatures below 50°C, but appears to be a parameter of inter-
est at temperatures above 50°C, which may make it an attractive
parameter for assessing high-temperature ablation.
1D temperature image using
thermal strain
52
20
50
30
48
46
40
44
50
42
40
60
38
70
0
13.4.3 thermal Strain Using Echo Shift
Temperature maps based on thermal strain to find the change in
temperature Δ
T
(
z
) along the propagation direction can be estimated
by differentiating Equation 13.1, the expression for echo shift
t
δ
(
z
):
58
10
20
30
°C
Lateral (mm)
FIGURE 13.7
(
See color insert.)
1D thermal strain temperature
image from axial echo shifts in turkey breast muscle during nonuni-
form heating using the fixture in Figure 13.5. Tissue was surrounded
by water at 37°C. At time = 0 seconds 65°C water was pumped through
the tissue center (black disk). Compare to the 3D CBE image in Figure
13.14 from a different specimen of turkey. Thermocouple readings in
planes adjacent to the image were used to scale the strain image to give
temperature estimates.
∂
∂
Tz
()
=
k
z
tz
(),
(13.6)
δ
where
k
is based on the change in SOS with temperature and the
thermal expansion coefficient α of the medium of interest. For
example, Maass-Moreno and coworkers investigated the ability
to predict temperature in HIFU therapy from echo shifts in tur-
key breast muscle.
58,89
They found that results were consistent
with their theoretical predictions. In investigations by Ebbini
and coworkers, tracking echo shifts from scattering volumes was
shown to be promising, as was the work of other investigators
looking at echo shift for temperature estimation.
44,60,88
Combining echo shift temperature estimation with elastog-
raphy can improve the quality of monitoring of thermal lesions
induced by focused ultrasound.
112
Furthermore, echo-shift estima-
tion has been implemented with a zero-crossing method for track-
ing temporal positions of echo to determine the difference between
A-mode signals gelatin phantoms before and after heating.
113
his
technique yielded as much as a seven-fold improvement in com-
putational efficiency compared to conventional cross-correlation
methods with similar temperature estimation results.
Sun and Ying have also found some success in being able to pre-
dict temperatures using time-gated echo shifts, but they acknowl-
edge the difficulty of using this method for general temperature
monitoring because prior knowledge of both SOS and thermal
expansion coefficients is necessary.
35
Obtaining
a priori
knowl-
edge of both SOS and thermal expansion coefficients is a formi-
dable problem for the
in vivo
case, as demonstrated by a quick look
at the early tissue-characterization literature, which shows that
SOS can vary greatly in different types of tissues. In fact the speed
change due to temperature in lipid tissue is opposite in direction
to the SOS change in aqueous tissue. These complications could
cause difficulties in determining temperature in the complicated
inhomogeneous tissues likely to be found in an
in vivo
situation.
Varghese, Zagzebski, and coworkers investigated the spatial
distribution of heating using echo shifts in studies that included
in
vivo
measurements
61
and more recently
in vivo
temperature esti-
mates during high-temperature ablation.
62
Temperature estimates
were obtained using cross-correlation methods described previ-
ously and in Equation 13.1. Resulting temperature maps were used
to display the initial temperature rise and to continuously update
a thermal map of the treated region that was simultaneously mon-
itored using thermosensors. Figure 13.7 shows a thermal strain
image in a specimen of turkey breast muscle we generated from
measurements taken with the fixture in Figure 13.5. The constant
k
in Equation 13.6 was estimated by comparison of the thermal
strain image to thermocouple readings.
13.4.4 Change in Backscattered Energy
In a search for an ultrasonic parameter that changed monotoni-
cally with temperature, we modeled the backscattered energy
from individual scatterers to an interrogating ultrasonic wave.
91
According to that model, the change in backscattered energy due
to temperature was primarily dependent on the changes in SOS
and density of the medium compared to their values in sub-wave-
length inhomogeneities (scatterers) within the medium. Our pre-
dicted change in backscattered energy (CBE) at any temperature
T
with respect to its value at some reference temperature
T
R
is
−α
−α
2()
2( )
R
Tx
Tx
=
α
α
()
()
T
T
η
η
()
()
T
T
[1
−
−
e
e
]
R
CBET
()
(13.7)
[1
]
R
