Consequence: ITD WA/A2F2

Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

Br. J. Cancer (1982) 45, Suppl.

V, 140

ACOUSTIC CAVITATION: A POSSIBLE CONSEQUENCE


OF BIOMEDICAL USES OF ULTRASOUND
R. E. APFEL

From the Department of Applied Mechanics, Yale University, P.O. Box 2159, New Haven,
CT 06520, U.S.A.
Summary.-Those concerned with acoustic cavitation often use different measures
and nomenclature to those who employ ultrasound for medical purposes. After
illustrating the connections between the two, acoustic cavitation phenomena are
divided into two classes: (1) relatively moderate amplitude changes in the bubble
size that occur during each acoustic cycle, as with rectified diffusion and resonant
bubble motion, and (2) rather dramatic changes in the bubble radius that occur in
one cycle. It is seen that pulse-echo diagnostic equipment can excite the dramatic
changes whereas continuous wave therapeutic equipment will excite the slower, but
no less important, changes. The ranges of the acoustic variables and material states
for which these phenomena are possible are quantified. It is shown that whereas
the concept of an ultrasonic (energy) dose may be appropriate for the effects of
acoustically induced heating or resonant bubble motion, it is inappropriate when
discussing the effects of the transient type of cavitation that can occur from short,
high amplitude acoustic pulses.

ACOUSTIC cavitation can be defined as


any observable activity involving a bubble
or population of bubbles stimulated into
motion by an acoustic field. Since this
cavitational activity can be viewed as a
dramatic concentration of acoustic energy
resulting in localized high stresses, temperatures and/or fluid velocities, its biological consequences should be understood
by those who are trying to either optimize
or minimize its effects.
For the purposes of this conference I
prefer not to attempt to review the many
review articles on the subject (e.g. Flynn,
1964; Neppiras, 1980; Coakley & Nyborg,
1978; Apfel, 1981a). Nor do I intend to
review the empirical evidence for the
effects of acoustic radiation on biological
tissues. The book by Hussey (1975)
reviews much of the work before 1975,
and many of my colleagues at this conference are far more qualified to discuss
the present status of work in this area.
Rather I prefer to relate and quantify the
parameters used in describing cavitational
activity with those parameters most often
associated with equipment used for bio-

medical applications of ultrasound. Then


I shall consider the range of these parameters for existing diagnostic and therapeutic equipment. And finally I shall
discuss the kinds of cavitational activity
that are possible in each case, and briefly
outline a methodology for assessing
thresholds for bioeffects.

Some parameters associated with cavitation,


with ultrasound equipment, and some relationships between the two
Tables I and II give parameters normally associated with cavitation phenomena
and with medical ultrasonic equipment,
respectively. What is immediately clear is
that the language used for one is rather
different to that used for the other. We
can find the useful relationships between
the two as follows:
At transducer:

ITD = W/7T(D2/4)
In focal region:

IFC= GIITD = WA/A2F2

141

ACOUSTIC CAVITATION

p (peak at focus) =
1P22 x 103 ipl/2(0); ip(O), in W/m2
or p estimated assuming uniform distribution across transducer face and in focal
region

I_'F
Tr x PRF
1i73x103 / WRxA \1/2
FA
rp x PRF
115f j 'xA 1/2 Win watts
F \-rpxPRF) ' for water (1)

=22pcx

Example 1: Pulse-echo system, nominal


frequency 2 MHz, 2 cycles per pulse
1 us),PRF= 1 kHz
(rplI
P=

6*46 x 107 Wl/2D/F, in Pa; It' in watts


20 4 W1/2D/F, in bars; TF in milliwatts
If W-= 10 mW and D/F= 0.3, then p(t)-~
19 bars=1-9 x 1.06Pa.
Example 2: Therapeutic system; f750 kHz, continuous wave, unfocused.
p= 1P73 x 103

195 x 10 W1/2 Ur in watts


D 'D in metres
For D = 0 06 m and TV= 2 watts,

(2)

p =46xlx4Pa-0*46 bars
Whereas the ultrasonic exposure (energyT
dose) is much lower for diagnostic equipment than for therapeutic equipment, the
peak acoustic pressures are much higher
in the diagnostic equipment! As we shall
see, this is an important consideration in
assessing the likelihood and type of
cavitation and its potential bioeffects.

Types of cavitation possible in different


types of sound fields
We can divide most cavitation phenomena into two categories: (1) small-tomoderate amplitude bubble motion in
which the change in the bubble radius in
one acoustic cycle is significantly less than

TABLE I. Some parameters associated with


acoustic cavitation
Acoustic variables
Instantaneous acoustic pressure
Acoustic frequency, periodl
Acoustic wavelength
If pulsed:
Pulse width
Pulse repetition frequency
T'hermodynamic state of materia l
Hydrostatic pressure
Temperature
Dissolved gas concentration
(fraction of saturation) or
partial pressure
Material properties
Density
Viscosity
Surface tensioni
Thermal conductivxity
Sound velocity
Compressibility
Specific heat
Vapour pressure (for liquids)
Bubble parameters
(assuming spherical)
Radius, (initial size)
Radius: max. size, min. size
Bubble interface velocity
Internal pressure
Internal temperature

Symbol

p(t)

f,

A
I-p

PRF
Po
Yo

Ci

PC"
p
/1

k
c

Pv

R(t), (Ko)
R max, Rmin
R
Pi

Y1i

the radius; (2) large amplitude bubble


motion in which the bubble grows in one
cycle by an amount equal to or greater
than the radius.
Small-to-moderate amplitudes.-In this
case the bubble may be growing or collapsing or just oscillating about an
equilibrium size. Non-condensible gas
within the bubble is responsible for the
relatively small amplitudes.
Growth is possible over many acoustic
cycles by a process called rectified diffusion, in which slightly more gas diffuses
into the bubble upon expansion than
leaves during contraction. Thus, if the
acoustic pressure is greater than some
threshold (see Lewin & Bjorno, 1981 and
Crum, 1980 for critiques) the bubble will
slowly grow. Richardson (1947) has shown
that at a radius of:
fJP O

Rr

Xr

=PoRr

[iX r(1

(3)

142

R. E. APFEL

TABLE II.-Some parameters associated with medical ultrasonic equipment*


Energy related measures
'Total transmitted acoustic power
Temporal average intensity near transducer averaged
Time
over area of active transducer elements
averages' Temporal average-spatial average at focus
Temporal average-spatial peak intensity measured
X at focal point or last axial maximum
Temporal average intensity in the focal plane, averaged over the
area in which the intensity exceeds 0*5 x I(O)
Spatial peak intensity averaged over the duration of the pulse
(temporal peak-spatial peak)
Instantaneous temporal peak-spatial peak intensity
Spatial measures
Radius of beam profile in focal plane for which the
temporal average intensity = I(0)/2
Focal length of transducer
Physical transducer diameter
Physical transducer area
Effective transducer diameter
r
Theor. intensity gain at focus
If spherically focusingI
Theor. pressure gain at focus
and D <F/2

Temporal measures
Nominal (centre) frequency
Pulse width, defined in terms of time between first and last zero
crossings in the wave form at which the instantaneous pressure
is i of the peak pressure
Pulse repetition period (rate)
* Some nomenclature taken from Carson et al., 1978.

the bubble will be in resonance (the


"gas" spring oscillating against the effective mass of the surrounding fluid, plus a
surface tension effect). Here K is a calcuable constant, which depends on heat
conduction and which is close to one for
micron-sized bubbles. For reference, resonant air bubbles in water have the following
radii for the given frequencies:
f, MHz
0 5 1
2
4
8
6 3-2 1-8 0*95 0*55
Br, /tm
It is noteworthy that in this frequency
range of medical ultrasonic equipment, the
resonance radius is comparable to the size
of biological cells, whereas it is a factor of
300-500 smaller than the wavelength of
sound at the frequency of resonance.
The threshold acoustic pressure amplitude for rectified diffusion in the low
megahertz range is on the order of tenths
of bars (1 bar= 10 Pa) for bubbles reasonably close to resonance size and for 100%
gas saturation. As the frequency goes up,
however, the threshold goes up; as an

Symbol
W

ITD
IF
I(O)
IO.5
Ip(O)

ip(O)
Ro.5
F
D
A = TD2/4

Derr

Gp

~~~~~~~~A
2
A
AF

f
Tp

PRP (PRF)

example, the threshold is the same for


bubbles approximately 50% of resonance
size at 0.5 MHz and about 80% of
resonance size at 2-5 MHz.
For continuous wave excitation, a peak
acoustic pressure of 105 Pa (1 bar) corresponds to a plane wave intensity of close
3 kW/m2, which is well within the range
of ultrasonic equipment used for physical
therapy. For a frequency of 750 kHz,
bubbles in aqueous solution in the radius
range of 2-5 ,um are likely to grow for
intensities of 3 kW/m2 or greater. The
rectified diffusion threshold will go up as
the fraction of the transducer on-time is
reduced, as surface tension works to
collapse the bubble.
Bubble growth by rectified diffusion is
not a particularly exciting process until
the bubble reaches resonance size, resulting in larger amplitude oscillations and a
variety of consequences in the medium. I
shall not dwell on this subject since
Professor Nyborg (1982) elaborates on
this topic elsewhere in these proceedings.

143

ACOUSTIC CAVITATION

It is worth noting, however, that if a


bubble is already near resonance size, for
the given frequency of excitation, then
the pressure amplitude for its excitation,
even at megahertz frequencies is rather
modest (< 0.5 x 105 Pa), with viscous
damping tending to predominate over
thermal and acoustic damping for the
conditions obtaining with biomedical uses
of ultrasound (Prosperetti, 1977). ter
Haar et al. (1982) present elsewhere in
these proceedings some evidence that the
effects of rectified diffusion at therapeutic
levels (, 104 W/m2) may occur in vivo.
Potential for cavitational damage from
pulsed ultrasound.-Biological damage
from short pulses of ultrasound can occur
if the growth of a bubble during one cycle
of tension (during the negative phase of
acoustic cycle) exceeds some defined
criteria (which will be discussed shortly).
The growth depends not only on the peak
acoustic pressure amplitude but also on
the time allotted for growth (which is less
than a half acoustic period). The lower the
acoustic frequency for a given pressure
amplitude, the greater the bubble growth
and the more violent the collapse.
Two factors which inhibit bubble growth
are the inertia and the viscosity of the
medium surrounding the bubble. Both
tend to impede the initial growth and for
both we can estimate a time constant
which can be thought of as a time for the
growth process to get started. If the sum
of the inertial start-up time (t1) and the
viscous start-up time (tv) is comparable
to or greater than a significant fraction
(say 0.2) of an acoustic period (r), then the
bubble growth will never really get started
before the phase of the acoustic cycle has
gone positive and the bubble begins to
decelerate prior to collapse. Restating this
criteria using expressions given by Apfel
(1981a, b), we have very rough estimates
of t, and tI summed as follows:

tv+tI= 4tp + 2Ri /P

<0-2,r

0-

(4)

where AP is the time-averaged pressure


while the driving pressure plus ambient

pressure is negative, =2(PA-Po)13, with


PA the peak acoustic pressure and Po is the
ambient pressure. Also , is the dynamic
viscosity, RI is the initial bubble radius, f
is the frequency (= l/ r), and p is the
density of the medium. In Fig. 1, we plot
with dots Equation (4) as Ri versus PA for
different frequencies. Here we choose a
blood plasma-like environment with ,u
0-002 N.s/m2, or about twice that of water.
We note that as the bubble gets bigger,
the inertia of the medium becomes considerable, and significant growth is seriously impeded; the possibility for a strong
cavitation event is correspondingly diminished.
What do we mean by strong cavitation
PEAK INTENSITY (W/m2)

xacp

._

-6

4 6 8 10 12 14 16 18 20 22 24 26 28 30
Peak Accustic Pressure Amplitude (BARS 105 Pa)

FIG. 1. Curves for the given frequency


delineating the initial bubble size or smaller
for which transient cavitation is possible
for a particular peak acoustic pressure.
The A's represent transient cavitation predictions not taking into account viscosity
or the initial inertial effects. The dots
represent the constraint implied by viscous
and initial inertial effects. A curve is drawn
through either the A or the dot, depending
on which corresponds to a lower radius for
a given pressure. The upper abscissa is
given in terms of peak acoustic intensity.
The assumed viscosity is 0-002 N.s/m2 or
about twice that of water.

144

R. E. APFEL

event? Here we are in a bit of a quandary.


We could define a transient cavitation
event by the criterion that the velocity of
the collapsing cavity approaches the
velocity of sound in the medium before
being decelerated by the gas trapped in
the cavity. This will occur, as shown by
Neppiras & Noltingk (1951), if the bubble
reaches a maximum size Rm equal to a
little more than twice the initial size
(2.3 to be more precise). This "collapse
velocity" criterion would be reasonable if
it were thought that the stresses and
velocities thus generated were the cause of
significant biological damage. This criterion, from Equation 18 of Apfel (1981b), is
plotted with A's for different frequencies
in Fig. 1. However, the criterion will not
be met if the startup criteria are not also
met. Therefore, for each frequency a solid
line is drawn through the dots and A's
which satisfy both criteria (transient
cavitation, given sufficient time for growth
in a fraction of an acoustic cycle).
Another possibility for a damage
criterion is based on energy. This "energy
criterion" is related to the absolute size
reached by the bubble as contrasted with
the "collapse velocity" criterion which
depends only on the maximum bubble size
relative to the initial size. It stands to
reason that the more potential energy
(- volume x pressure difference across bubble) stored in the fluid at its maximum
size, the greater the kinetic (and eventually thermal) energy given back upon
the bubble's collapse.
One approach in considering the possibility of cavitational damage in biological tissues in a pulsed ultrasonic field
might be to say that the "collapse
velocity" criterion is met and, in addition,
the maximum bubble radius equals or
exceeds some absolute size criterion. For
individual cell damage it does not seem
unreasonable initially to choose this absolute size as that of a cell. For example, let
us take the effective radius of a cell to be
about 5 Kim; then we hypothesize that
there will be damage to the cell if both the
"collapse velocity" criterion is met and

the bubble grows to a radius of at least


5 [m. Apfel (1981b) gives an approximate
expression (his Equation 16) which relates
the maximum size a bubble will reach to
the acoustic pressure and frequency. This
can be written as follows:

fMHz R14'm ,3

(p

2) [I1+ 2(P- 1)/3]1/3

(5)

p1'2

where fMHz is the frequency in MHz,


and
Rp,m is the maximum radius in ,um,MHz,
P =PA/PO. For R = 5 pm, and f= 2

p_6.

In Fig. 2, we show the 2 MHz case from


Fig. 1 with a vertical line at PA/PO =6,
corresponding to a 5 ,um maximum radius
criterion. The cross-hatched region represents the radius-pressure combinations for
which both velocity and energy criteria
are met. Note that PA= 6 105 Pa corresponds to a 100 W/m2 time averaged
intensity assuming 2 cycles per pulse and
a pulse repetition frequency of 1 kHz.
x

(W/m2) Assuming 2 Cycles


kHz Pulse Repetition Frequency

Time-Averaged Intensity
per Pulse and

102

[T

X 102

103 1.5 X

103

Peak Acoustic Pressure Amplitude (BARS

i05Pa)

FIG. 2.-Transient cavitation possibilities at


2 MHz, taken from Fig. 1. Also drawn is a
vertical line at pA/PO = 6. The hatched
region gives the initial bubble radiuspressure combinations for which both
transient cavitation is possible and the
bubble will reach a radius of at least 5 ,um.
The upper abscissa corresponds to the
time-averaged intensity assuming two
cycles per pulse and a pulse repetition
frequency of 1 kHz.

145

ACOUSTIC CAVITATION

Although there appears to be no lower


size limit to the hatched region, there is
a lower, surface tension-dependent, nucleation threshold for the initiation of the
growth of a gas bubble, which Blake
(1949) formulated and which can also be
found in Apfel (1981a). For acoustic
pressures in excess of 6 x 105 Pa, however,
the threshold radius is below 0-1 pm, and
therefore does not appear on the graph.
It is entirely reasonable at this stage of
the discussion to challenge the suggested
criteria, because the macroscopic effects
normally associated with transient cavitation usually result from multiple bubble
effects. That is, in a continuous wave field
one small bubble collapsing will shatter,
producing seed bubbles that will grow
during the later cycles. This collection of
bubbles will go through this process often,
as the photographs of Crum (1972) suggest, until the collection of bubbles is of
resonant size, at which point radiation
forces push the collection away from
regions of high acoustic pressure.
For the diagnostic ultrasound case, only
one growth and collapse phase is likely
and the total energy is extremely small.
It is simple to show that a bubble that
has grown substantially to a 5 ,um radius
has a potential energy of about 300 MeV
(million electron volts). Nevertheless, the
scale we are talking about is of the order
of the size of a cell. What must be determined, therefore, is the appropriate scale
for damage that results in acceptable
risks. The vertical line in Fig. 2 can be

thought of as a line-cursor which can be


moved horizontally, depending on our best
estimate of reasonable damage criteria.
A few additional comments on high
amplitude growth.-Figs 1 and 2 refer to a
viscosity of 0-002 N. s/M2. Thresholds for
bubble activity in tissues of much higher
viscosity will be even higher, making it
likely that the only concern for damage
due to cavitation will be with aqueous
environments. For instance, for a viscosity
of 002 N s/m2 (a factor of 10 greater than
considered in Figs 1 and 2) a pressure
greater than 1.5 x 106 Pa would be required for the possibility of transient
cavitation.
Our discussion has also left out absorption, which will tend to diminish cavitation, and focusing and standing waves,
which may tend to enhance it.
CONCLUSION

Table III summarizes the cavitational


effects that are possible in aqueous media
from biomedical ultrasound ranging from
pulsed to continuous and from moderate
to high acoustic pressure amplitudes.
Whereas viscous effects can eliminate the
possibility of transient cavitation from
pulse-echo ultrasound, it is probably less
important for rectified diffusion. And
whereas the concept of energy dose may
be relevant when discussing heating due
to ultrasound, or the effects of resonant
bubble motion, it is less appropriate for
transient cavitation produced by short

TABLE III.-Types of cavitation posu&ble for seed bu,bble in aqueous environment


Continuous wave
(or significant fraction of on-time)

Pulsed

Moderate pressure ampl;


Therapeutic and doppler
devices
(PA< 1-0 atm)

1. Bubble growth by rectified diffusion


2. Mechanical activity produced by
resonant bubble oscillations

none. for short duty cycle

Higher pressure ampl.

1. Dynamic cavitation
2. Bubble break-up and new growth of
several bubbles-collective
cavitational effects
3. Significant mechanical activity

Significant bubble growth and


collapse possible in one acoustic
cycle if pressure is great enough
and if frequency is low enough
(see Figs 1 and 2). Time-averaged
spatial peak intensities can be
as low as 100 W/m2

(PA > 3 atm)

146

R. E. APFEL

bursts of high pressure amplitude ultrasound.


For the future, the question of the
existence of seed bubbles in mammalian
tissues requires a definitive answer. Moreover, experimentalists seeking to determine whether or not there are bioeffects
from ultrasound must take care to know
the acoustic field, which means choosing
equipment and geometrical configurations
which are most conducive to knowing the
instantaneous pressure as a function of
position, as well as the various integrated
energy measures, the time histories,
temperature, and the state of the medium
(gas saturation, solution properties, etc.).
This work is supported in part by the Physics
Program of the U.S. Office of Naval Research. I
thank E. A. Neppiras for his advice on this general
subject.

REFERENCES
APFEL, R. E. (1981a) Acoustic cavitation. In
Ultrasonics. Ed. P. Edmonds. Methods of Experimental Physics Series, Ed. Marton. New York:
Academic Press.
APFEL, R. E. (1981b) Acoustic cavitation prediction.
J. Acoust. Soc. Am., 69, 1624.
BLAKE, F. G. (1949) Tech. memo 12, Acoustic
Research Lab., Harvard University.
CARSON, P. L., FISCHELLA, P. R. & OUGHTON, T. V.
(1978) Ultrasonic power and intensities produced

by diagnostic ultrasoumd equipment. Ultrasound


Med. Biol., 3, 341.
COAKLEY, W. T. & NYBORG, W. L. (1978) Cavitation: dynamics of gas bubbles; applications. In
Ultrasound: Its Applications in Medicine and
Biology. Ed. F. J. Fry. New York: Elsevier.
CRUM, L. A. (1972) Velocity of transient cavities in
an acoustic stationary wave. J. Acoust. Soc. Am.,
52, 294.
CRUM, L. A. (1980) Measurement of the growth of
air bubbles by rectified diffusion. J. Acoust. Soc.
Am., 68, 203.
FLYNN, H. (1964) Physics of acoustic cavitation in
liquids. In Physical Acoustics. Ed. W. P. Mason.
New York: Academic Press. p. 57.
HuSSEY, M. (1975) Diagnostic Uttrasound. Glasgow:
Blackie.
LEWIN, P. A. & BJ0RNo, L. (1981) Acoustic pressure amplitude thresholds for rectified diffusion
in gaseous microbubbles in biological tissue. J.
Acoust. Soc. Am., 69, 846.
NEPPIRAS, E. A. & NOLTINGK, B. E. (1951) Cavitation produced by ultrasonics: theoretical conditions for the onset of cavitation. Proc. Phys. Soc.,
B64, 1032.
NEPPIRAS, E. A. (1980) Acoustic cavitation, Phys.
Rep., 61, 160.
NYBORG, W. L. (1982) Ultrasonic microstreaming
and related phenomena. Br. J. Cancer, 45, Suppl.
V, 156.
PROSPERETTI, A. (1977) Thermal effects and damping mechanisms in the forced radial oscillations of
gas bubbles in liquids. J. Acoust. Soc. Am., 61, 17.
RICHARDSON, J. M. (1947) As quoted in footnote 14
of Briggs, H. B., Johnson, J. B. & Mason, W. P.
(1947) Properties of liquids at high sound pressure.
J. Acoust. Soc. Am., 19, 664.
TER HAAR, G., DANIELS, S., EASTAUGH, K. C. &
HILL, C. R. (1982) Ultrasonically induced cavitation in vivo. Br. J. Cancer, 45, Suppl. V, 151.

You might also like