ACOUSTICAL CHARACTERIZATIONOF ULTRASONIC SURGICAL DEVICES

Mark E. Schaferl and Alan Broadwin2

1Sonic Technologies, Inc., 2935 Byberry Rd. Hatboro, PA 19040 U.S.A.

2Ainslie, 27 Pine Ridge Rd. Larchmont, NY 10538 U.S.A.

ABSTRACT Ultrasonic surgical instruments operating cates that acoustic activity such as cavitation is involved.
from 20 to 60 k H z have typically been characterized only These devices have traditionally been measured using non-
by their operating frequency, cutting tip excursion, and acoustic parameters, i.e. displacement, frequency, and
input electrical power. This paper presents acoustical input electrical power. Acoustic parameters would include
measurement methodologies which more completely acoustic pressure (in units of Pascals, or N/m2) and its
characterize device output. The methods are divided into distribution in time (frequency) and space (position rela-
two frequency regimes: frequencies below 100 kHz whch tive to the tip), acoustic source strength (to be discussed
correspond to the tip driving frequency, and frequencies in detail later, in units of m3/s), and acoustic power (in
above 1 MHz whch corresponds to the cavitation devel- units of Watts). Only limited prior work has been done on
oped at the tip. Low frequency radiation is modelled as the acoustical characterization of these devices [I], whch
an acoustic dipole; high frequency acoustic radiation is noted the presence of cavitation noise, but did not attempt
caused by cavitation bubble collapse at the tip, generating to quantify it.
spherically diverging shock waves. This measurement
method incorporates the energy in a single shock wave THEORETICAL MODELLING Basic Low Freauency
event, the number of events per unit time, and the spheri- Monopole Model The theoretical acoustic model starts
cal divergence of the wave. Combining the two measure- with the monopole radiator, radiating as a point source of
ment techniques shows the growth of the total radiated spherically expanding waves. After some manipulation,
energy as a function of excursion, and a commensurate the most straightforward expression for the pressure p at a
shift from low frequency energy to cavitational energy. distance r from a monopole source pulsating at a frequen-

INTRODUCTION The overall goal of this work was to where po and c are the density and sound speed of the
develop reliable and meaningful measurement procedures medium, k is the wavenumber (w/c), and Qs is the equiv-
for use with ultrasonic surgical equipment (e.g. neurosur- alent source strength. The acoustic source strength, or
gical and phacoemulsifier devices). The primary rationale "volume velocity," for a small vibrating body is the
is to measure relative performance between surgical tips product of the radiating surface area times the normal
of different types, and to perhaps predict surgical perfor- velocity of that surface. Thus the source strength can be
mance. Manufacturing quality control is another concern found from two n times the product of tip area A, tip
to be addressed, as well as the development of interna- vibrational amplitude 6 (one half the excursion), and
tional standards for devices of this type (i.e. the IEC). driving (radian) frequency w. Substituting these into equa-
tion l gives an expression for the pressure as a function of
Ultrasonic surgical instruments typically use hollow stain- distance from a monopole source:
less steel or titanium tubes, which are vibrated longitudi- jP@ 2~6e+'
nally by a high power ultrasound source. Newer applica- P= 47~y (2)
tions include catheter wire systems for arterial plaque The acoustic power can, be expressed
_, in terms of the
disruption, as well as blade or scalpel type devices. The source strength Qs: w = [ g ) ( Q ]
frequency of operation ranges from 20kHz to 60kHz, (3)
depending upon the application. The displacement of the The power radiated from the tip (as a monopole) is thus
needle tip "fragments" the tissue which is in immediate proportional to A2 x 82 x f4. When the tip is immersed
contact with it. It is still not clear at this time the exact into a water tank io a depth less than 5-6 wavelengths, the
mechanism of fragmentation, but the leading theory indi- result is an effective dipole source. Measurements of

1051-0117/94/0000-1903$4.00 0 1994 IEEE 1994 ULTRASONICS SYMPOSIUM - 1903

ultrasonic surgical catheters which did not produce a energy of a single pulse, assuming that the collapsing
dipole geometry were found to agree with Equation (2). bubble is a monopole radiator. Thus the area factor
becomes 4m2, and with the hydrophone at 1 cm from the
The implications of this geometry and source assumption tip, the factor is approximately 5 x (W-spa2).
include: a 1/R dependence of the pressure radiated from
the tip, a dipole-like radiation pattern (pressure as a func- The next step is to sum the individual shock waves over
tion of O), and distortions of the radiation pattern for rela- some measurement time interval. A computer program
tively small values of R. All these assumptions have been was used to sample time waveforms which contain a
borne out by the comparison of simulation and experi- number of shock waves, filter out the low frequency
ment. For the simulation work presented later, the pres- components (<46kHz), cutoff all pressure values less than
sure at any observation point was found using Equation lOWa (to reduce noise), and square the resultant wave-
(2) as applied to a dipole geometry. form. Next, the computer sums up the signals in the wave-
form and divides by the observation time This can be
For the experimental determination of source strength, the expressed as: ~ P S ( 5 1 o-”)
(6)
analysis was simplified to the case of pressure along the Observation Time
tip axis direction. Further, a “far-field” simplification was Combining the energy in a single shock wave pulse and
made that R>d, which was generally true for the immer- the rate of the pulses, results in a measurement of the
sion depths and hydrophone positions used. Under those power involved in cavitation activity. Both portions are
conditions, the “on-axis” pressure can be expressed as: equally important, since cavitation appears to occur at
(4) different rates as a function of driving energy (tip excur-
sion). At some excursions, cavitation was observed every
This equation can be rewritten to find the effective source cycle, and at other excursions, it was observed every other
strength from a pressure measurement taken at a distance cycle.
r, with a source separation of d (note that d is twice the
To summarize the assumptions used: I ) the signal from
the high frequency hydrophone is a representative pres-
Cavitatioq It is clear fiom even casuai’observation that sure-time waveform from the bubble collapse; 2) the
ultrasonic surgical tips generate bubbles due to cavitation. signal may be analysed to find the pulse-intensity integral;
This produces an additional source of acoustic radiation: 3) the shock wave shape is relatively independent of abso-
secondary pressure waves generated by cavitation bubble lute peak pressure, so that a normalization factor can be
collapse. When these cavitation bubbles collapse, they used to convert any Pc value into a pulse energy for that
emit a wideband acoustic shock wave pulse, with a center excursion level; 4) the shock wavefront is essentially a
frequency much higher than the driving frequency of the spherical wave; and 5) the summation procedure is an
probe. In addition, the cavitation is not completely coher- adequate way of characterizing the overall number and
ent with the dnving signal, in that the source of the cavita- amplitudes of the individual shock wavefronts.
tion event is the collapse of a randomly positioned bubble
in the vicinity of the radiating tip. Finally, it is not possi- MATERIALS & METHODS The measurement system
ble to measure too closely to the tip of the probe, since the used was comprised of calibrated hydrophone probes, a
cavitation action may damage the hydrophone. Therefore, PC/386 compatible computer, a Tektronix 2440 Digital
the cavitation signal is measured at some reference dis- Oscilloscope, a water tank (22°C deionized water), and a
tance from the tip. stepper motor driven micromanipulating system. The
measurement tank was 2‘ x 2‘ x 4‘ long, with the central 2‘
Cavitation energy quantification is a multi-step process. x 2’ x 2’ volume lined with 1/2 inch thick SorbothaneTM.
First, the energy of a single shock wave is analysed to Validations tests showed that this small tank, when lined
produce a normalization factor between peak compres- with acoustically absorbing material, was sufficiently
sional pressure and the energy in a single pulse, using the anechoic to permit pressure measurements comparable to
Pulse Intensity Integral (PII). This assumes that any other a much larger tank (a 5 x 10 x 3 meter deep outdoor
pulse resulting from another bubble at the same amplitude swimming pool). In addition to standard rectilinear posi-
setting has essentially the same pulse shape, only a differ- tioning, a specialized rotary fixture was developed to scan
ent pressure amplitude. A typical normalization factor the hydrophone about the tip, in order to measure the
was 4 x in the somewhat unconventional units of acoustic field distribution as a function of angle. The
W-s/cm2-Pa2. The next normalization gives the source positioning resolution and repeatability was 0.03 175 mm

1904 - 1994 ULTRASONICS SYMPOSIUM

in the X and Y directions (lateral with respect to the at the h/4 point. The discrepancy at the second peak is
acoustic axis), 0.0635 mm in the Z direction (axial with probably due to shielding of the secondary (image) source
respect to the acoustic axis), and 0.05’ in the 0 direction by the probe itself. The conclusion it that the h/4 point
(rotation about the surface point). The oscilloscope and provides the position of both maximum pressure and
stepper motor controller were interfaced to the PC/AT minimum error from positional variation. The next test
computer via an IEEE-488 bus for capture and storage of involved scanning around the tip immersion point in an
the hydrophone signals, and for automated positioning of angular motion. Again, this was both simulated and tested
the hydrophone. The system is similar to the one used for experimentally, and a typical comparison is shown in
testing diagnostic ultrasound equipment [2]. Figure 2 for distances of 2.5 and 5.0 cm from a 23kHz tip.

A Briiel and K j m Type 8103 miniature piezoceramic In order to estimate the low frequency source strength, the
hydrophone, with a frequency response from 100 Hz to pressure as a function of distance from the tip was meas-
80 kHz, was used for the low frequency measurements. ured, and analysed using the equations given earlier. The
For the high frequency (cavitation) measurements, a results fit well to the expected 1/R dependence, although
PVDF type needle hydrophone manufactured by the there were consistent anomalies which were found (une-
Danish Institute of Biomedical Engineering (DIBE) was xplained, repeatable deviations from the l/R curve). This
used. This hydrophone has essentially uniform frequency scanning approach is very useful, because the least-
response from 1 to 15 MHz. squares fit through the measured data can be used to make
an estimate of pressure at any convenient distance from
Additional items included a frequency meter and an elec- the source, even those distances which are not conducive
trical wattmeter attached to the surgical system unit, and a to direct measurement (such as within 1 mm of the tip).
peristaltic pump to circulate water through the surgical
tips. A triggering signal was taken from an available test- In order to measure cavitation, the PVDF needle hydro-
point on the drive electronics circuit board. The measure- phone was positioned on-axis, at a distance of 1 cm from
ment system was also able to monitor the trigger voltage the tip. The measurement oscilloscope was set to capture
and determine the excursion level of the tip, based on a high speed events, using the cavitation signal itself as a
separate optical calibration. trigger. The signal was averaged to improve the signal-to-
noise ratio. At each tip excursion setting, the cavitation
Selection of TiDs A number of surgical tips operating at waveform (if it existed) was captured, transferred to the
23kHz were built to allow intercomparison between dif- computer, and analysed to produce the source energy
ferent tip parameters, such as area and O.D. Hollow tips normalization factor. The normalization factor was com-
were made with different diameters and wall thicknesses; bined with a count of the number of cavitation events to
a number of “Solid” (flat end) tips were also built, find a time-average energy. Using the same set-up, the
because these would have fairly well defined radiating oscilloscope setting is changed to sample several cycles of
surface areas. The solid tips had the following 0.D.s: 1, the low-frequency waveform. This time, the analysis
1.5,2., and 2.5 mm (corresponding to areas of 0.79, 1.77, included filtering (to remove the low frequency compon-
3.14, and 4.91 mm2); the hollow tip sizes were: 1.9mm ent of the wave), thresholding (to reduce noise), and
“Micro Tip”, 2.54mm “Thinwall”, 2.54mm “Standard”, squaring and summation. The resultant data was multi-
and 2.77mm “Heavywall” (with corresponding areas of plied by the corresponding single-event energy normaliza-
0.82, 1.33, 1.83, and 2.77 mm2). A single tip operating at tion factor. Figure 3 shows the predicted (theoretical) and
37kHz was also tested (1.9 mm O.D, 0.82 mm2 Area). measured source strength for both the hollow and solid
tips as a function of tip area (units of m2, Q~ is in m3/s).
MEASUREMENT PROCEDURES & RESULTS The The solid lines are the theoretical Qs predictions and the
first test was performed to determine tip immersion depth dashed lines and individual symbols are the measured
which produces the highest pressure levels in the forward values. Theory and measurement for the same tip are
direction (generally the point of h/4 immersion). This denoted by similar symbols. Note that cavitation meas-
measurement was also simulated using the model equa- urements are included in this data. The data follows the
tions tailored to the experiemental conditions. Figure l theory fairly well, except that the solid tips tend to deviate
shows the comparison of simulation and experiment for a significantly at higher excursions. The final figure exam-
23kI-I~tip. In this case, the levels have been normalized, ines the effect of including cavitation in the measurement
since the simulation program used did not account for and the effect of changing frequency. The primary means
actual source pressure levels. The normalization was done of comparison is the 37kHz tip and the “thinwall” tip.

1994 ULTRASONICS SYMPOSIUM - 1905

Although these two tips have different areas (1.33 vs 0.82 and measurement of ultrasonic surrgical instrumentation.
mm2), outside diameters (2.54 vs 1.9mm), and frequen-
cies (23 vs 37kHz), they have almost the same predicted REFERENCES
source strength for a given displacement. If the source
strength theory is correct, then these two tips should have 1. Beissner, K. "Schallfelduntersuchungen an ophthal-
similar measured source strength levels. Figure 4 shows mologischchirurgishen Ultrashchallferaten," Fortchr. d.
the theoretical and measured Qs for the thinwall and Akustik, DAGA 80, VDE-Verlag, Berlin 1980, pp. 567-
37kHz tips. There is only one "theory" line, since both 570.
tips have essentially the same source strength. The data 2. Schafer, M.E. and Lewin, P.A. "A Computerized
shows that up to lOOpm displacement, the two operate System for Measuring the Acoustic Output from
nearly identically. The figure also shows the contribution Diagnostic Ultrasound Equipment," IEEE Transactions on
of cavitational energy analysis to the data. The dashed Ultrasonics, Ferroelectrics,and Frequency Control, UFFC-
lines represent the results without cavitation analysis. 35(2), pp. 102-109, 1988.

CONCLUSIONS In summary, the methods described,

based on source strength and cavitation analysis, have
shown good correlation between theory and experiment,
and provide addtional tools for the design, development,

1.2
Expcrimall

10.'
Tip Area (m2)

Figure 3. Source strength versus tip area, all tips.

0.0 ! I
0 20 40 60 80
Immersion (mm)

Figure 1. Pressure versus tip immersion, simulation and

experiment.

s2 10
E:
v:

1' I
10 100 loo0
Displacement (km)
no

Figure 2. Pressure versus angular position at 2.5 and Figure 4.Source strength versus displacement, 23kHz
5.0cm, simulation and experiment. "Thinwall" and 37kHz tips, with & without cavitation.

1906 - 1994 ULTRASONICS SYMPOSIUM