PULSE-ECHO FIELD DISTRIBUTION MEASUREMENT TECHNIQUE FOR HIGH-FREQUENCY ULTRASOUND SOURCES Kay hum and William D. O‘Brien, Jr. Bioacoustics Research Laboratory, Beckman Institutefor Advanced Science and Technology, Department of Electrical and Computer Engineering, University of Illinois, 405North Mathews Avenue, Urbana, IL 61801 (Current address for KR Institute of Medical Physics and Biophysics, Martin-Luther-University Halle-Wittenberg, D-06097 HalldSaale, Germany) Abstract - A simple technique for the determina- As long as the wire is sufficient small (h’< I, tion of the spatial and temporal transmit-receive where k is the wave number and a ’ is the radius fielddistribution of spherically focused high- of the wire) resonance modes are not excited by frequency transducers is described. In this study an incident wave and the wave scattered by the tungsten wires were used as echo-targets. Based cylinder has the form [l?] on the scattering of sound on a rigid cylinder the transmit-receivefield projection ofspherical sources measured with a wire target was com- pared with both, theoretical pressure distributions and hydrophone measurements in the frequency rangefrom 3 to 17 MHz. It wasdemonstrated that both techniques yielded comparable results for thelow-frequency transducer, whereas only the results of the wire target technique were also where K are the compressibilities and p are the in agreement with theoryfor the higher frequency densities, respectively. transducers. The lateral acoustic pressure distribution in the focal plane of a spherical focusing source can be INTRODUCTION described by [lo, 1 l] The use of spherically-focused, high-frequency transducers is widespreadin many high-resolu- tionultrasonic imaging applications such as in 2JI (&l medical diagnosis and material investigations. ’ (2) In this contributiona simple transmit-receive lgl= field projection technique is proposed to estimate ROC spatial and temporal field quantities of focused high-frequency transducers. Tungsten wires with where p(r) is the peak acoustic pressure as a diameters less than the acoustic wavelength were function of the off-axis lateral distance r, p(0) is used as echo-targets. Using a smalldimeter wire the on-axis peak acoustic pressure at z = ROC, k insteadof apoint-like target enables one to is the wave number, a is the radius of the trans- choose a target size (wire diameter) smaller than ducerand ROC is thetransducer’s radius of theacoustic wavelength, even for bigher curvature. J,(x) is the Bessel function of the first fiequencies. Tbis provides good spatial (in axial kind of order one. and lateral scan direction)andtemporal The axial pressure distribution of aspherical resolution, while the received signal amplitude is focusing source is described by[ 10, 1l] still sufficient due to the larger target area which is orientedperpendicular to the lateral scan directionandthe beam axis, respectively. Additionally, as long as an axial-symmetric field distributioncan be assumed only two scandi- rections are necessary for a spatial field projec- where p(z) is thepeak acoustic pressure as a tion. function of the axial distancez. 0-7803-4153-8/97/$10.00 0 1997 IEEE 1997 IEEE ULTRASONICS SYMPOSIUM - 1747 For a pulse-echo wire target, the line integral of and pulser/receiver were connected to a GPIB- the point-to-point distribution is obtained, that is board and controlled by a 486-66 PC. The time the pressure projectionf,(x,.z) : window at the oscilloscope was moved with every axial scan movement to maintain a high sampling f,(.,Z, = IP(X,Y,.z)dy> (4) rate (500 Msls) with a low number of sample points. Each 512-point W-signal was stored to where the wire is parallel to the y axis. The ge- the hard disk and transferred to a SUN Sparc 20 ometry dictates that the beam must be symmetric workstation for off-line processing. For about the beam axis and have relatively weak side comparison the transmitted field distribution was lobes in order to properly characterize the field's measured with a calibrated PVDF bilaminar dimensions near the beam axis. These conditions membrane hydrophone with an effective diameter are satisfied for strongly focused beams from a of 0.785+0.007 mm (Model 804, Sonic spherical transducer. Since the pulse-echo field is Industries, Hatboro, PA). All computations were composed of the multiplication of the transmitted performedwith MATLAB@ (The Mathworks, and received fields, equations (2) to (4) have to be Inc., Natick, MA). squared for the distributions of the transmit- 2D - field projections were obmned from the receive field. high frequency probes using a 25 pm tungsten wire (Figure 2). In the focal plane the lateral EXPERIMENTS field distribution wasalso assessed from all Tungsten wire targets with different diameters transducers using different wire diameters and a (25, 37, 63, 80 pm; California Fine Wire Com- hydrophone, respectively. All spatial graphs are pany, Grover City, CA) were placed in a tank plotted as the transmit-receive distribution of the pulse intensity integral (HZ)in dB. The frequency filled with distilled, degassed water (S 20°C) and oriented normal to the sound beamdirection spectra obtained with a 25 pm wire at the focal (Figure l). The targets were scanned across the pointwere used to estimate the theoretical acoustic field using a computer-controlled micro- lateral/axial distributions with eq. (2), (3) and (4). precision positioning system (Daedal Inc., In order to simulate a transmit-receive field the Hamson City, PA) with positional accuracy of calculated as well as the hydrophone determined about 2 pm, The grid size spacings in the lateral distributions were squared (Fig. 3). and axial directions were 25 pm and 50 pm. The Pulse duration ( q.20&,). bandwidth (AB, center transducers were excited by a 300 V mono-cycle frequency v,) and fractional bandwidth were de- pulse produceda by computer-controlled termined from the W-signal atthe focal point. pulser/receiver (Model 5800; Panametrics, Waltham, MA). The received signal was ampli- RESULTS fied (20 dB) and band-pass filtered (1-35 MHz) Figure 2 shows a transmit - receive field pro- by the pulsedreceiver. For all measurements, the jection of the 20 MHz -probe obtained with a 25 signal was displayed (500 M&) on a digitizing pm wire. Each contour line represents a 3 dB oscilloscope (Model 11401; Tektronix) with a 10- decrease of the PII value. The true focal length bit resolution. Positioning system, oscilloscope was determined from the on-axis time-of-flight of the received signal at the maximum PII location. The measured focal length (F), depth of focus (Fz)and beamwidth at the focal plane are comparable to the calculated values (Figure 2 and Table I). Pulse duration, bandwidth and center frequency were obtained from the true focal point RF signal using a 25-pm wire for the high frequency probes and a 80-pm wire for the 3 MHz probe, respectively (Figure 3). It was found that the 35- MHz (lY9 MIb.,, WhYdWDonC) pulsedreceiver used in this study appeared to attenuate thehigher frequency components Fig. 1. Block diagram of the main system components 1748 - 1997 IEEE ULTRASONICS SYMPOSIUM demonstrates thatthe hydrophone's effective diameter is dominant overthat of thefield's actual lateral distribution when the target size is larger than significant spatialchanges of the field's pressure distribution. Different wire diametcrs did not have an appre- ciable effect on the measured beam width (man agreements to within 6.8. 6.6 and 2.9% for thC 3-: 15- and20 MHz transducers, respectively). All measured lateral beam distributions were in the I ~l.0" " " " ' range between the calculated wire projections and 10 ,: I, I4 I5 Axd I)tatansslmm> the point-to-point distributions (near thc beam Fig. 2. Contourplot of thespatial intensity axis),and therefore, the experimental ob- distribution (PII in dB) of the 20 MHz transducer servations supportthe sugqestion that a wire field. Each contour line indicates a decrease of the diameter less than the acoustic wavelength is PII of 3 dB sufficient to obtain the significant field dimen- which resulted in the estimated center frequencies sions for spherically focused fields. of thetwo higher frequency probes being considerably lower than the manufacturer-stated CONCI.USION frequencies, that is, 13.05 MHz for the 15-MHz The results demonstrate thatthe wire-target transducer and 17.33 MHz for the 20-MHz technique is a simple and powerful measurement transducer. A comparison of wire determined procedure to determine the spatial and temporal lateral/axial field distributions (wire (M)) with transmit-receive acoustic field quantities from theoretical results (point (TH) and wire (TH)) high-frequency sources when an appropriately shows very good agreement near the focal region small effective hydrophone diameter is not avail- (PI1 > -10 dB). However, squaredthe able. For the low-frequency case investigated hydrophone determined lateral distribution herein, the hydrophone and wire-target techniques (hydrophone (M)) in the focal plane is in yielded comparable rcsults. For thetwo high-- agreement with theoretical/wire determined frequency cases, the wire-target results for the 3 MHz probe only. This Pulse-EchoResponse Power Density Lateral lntensity Anal Intensity Spectrum Distribution Distribution Panametrics V3 17 (20 MHz) Panametrics V3 19 (15 MHL) Panametrics V3680 (3 MHz) Fig. 3. Pulse-echo response, power density spectrum, lateral and axial intensit? distributions. All graphs arc plotted as transmit-receive distributions.