Time Reversal Acoustic Focusing With A Catheter Balloon

Ultrasound in Med. & Biol., Vol. 36, No. 1, pp. 86–94, 2010 Copyright Ó 2010 World Federation for Ultrasound in Medicine & Biology Printed in the USA. All rights reserved 0301-5629/10/$–see front matter

doi:10.1016/j.ultrasmedbio.2009.08.004

d

Original Contribution TIME REVERSAL ACOUSTIC FOCUSING WITH A CATHETER BALLOON YEGOR D. SINELNIKOV,* ALEXANDER M. SUTIN,y ANDREY Y. VEDERNIKOV,z and ARMEN P. SARVAZYANy * ProRhythm, Inc., Ronkonkoma, NY, USA; y Artann Laboratories, West Trenton, NJ, USA; and z Siemens Healthcare Sector, Moscow, Russia (Received 1 December 2008; revised 22 June 2009; in final form 5 August 2009)

Abstract—The ability to deliver configurable myocardial lesions was noted as a critical factor to the success of atrial fibrillation (AF) treatment. This article considers the implementation of time reversal acoustics (TRA) principles for ultrasound focusing using an AF cardiac catheter developed for pulmonary vein isolation. Experiments conducted with a single transmitting channel demonstrated that a catheter balloon could be used as an acoustic reverberator to enable focusing and steering of ultrasound short pulses in the TRA mode. The spatial effectiveness of the TRA focusing was improved using a catheter balloon of irregular, asymmetric shape and using a binary mode of ultrasound radiation. The experiments demonstrated the ability of steering the focal point over several millimeters without degradation of the focusing quality. An ability of the TRA mode to produce suitable therapeutic application focusing of long continuous ultrasonic signals was characterized in a theoretical model. (E-mail: [email protected]) Ó 2010 World Federation for Ultrasound in Medicine & Biology. Key Words: High-intensity focused ultrasound, Time reversal acoustics, Catheter, Atrial fibrillation.

INTRODUCTION

focusing in the TRA system. The excellent focusing ability of TRA has been used in various biomedical applications such as focusing through the skull and ribs (Fink 1997; Fink et al. 2003), where a multi-element array was used. Effective TRA focusing can be constructed using just one or very few transducers attached to an acoustic reverberator, where numerous reflections from the reverberator walls form a multi-element virtual phased array (Montaldo et al. 2004b; Anderson et al. 2008). TRA focusing systems, composed of several transducers attached to an acoustic reverberator, were also used in several nonmedical applications such as nondestructive testing of materials (Sutin and Johnson 2005) and land mine detection (Sutin et al. 2006). The TRA focusing systems described in the literature are based on the use of solid and liquid reverberators (Fink 2008; Sinelnikov et al. 2006; Sutin and Sarvazyan 2003; Fillinger et al. 2007; Quieffin et al. 2004), with one or several transmitters glued to reverberator facets. A single transmitter is capable of focusing a broad frequency band signal with time reversal, whereas it is not possible to focus a long monochromatic signal with single-channel time reversal (Derode et al. 2002). The narrow band signals typically require the construction of a time reversal mirror that consists of a large number of elements and time reversal channels (Fink 2008; Tanter et al. 2007).

Ultrasonic energy focusing is used in many medical therapeutic applications. Conventional means for acoustic focusing such as concave transducers, mirrors, lenses, and phased arrays work well in homogeneous media, but it is challenging to apply them in inhomogeneous media. For phased arrays, there are various methods of phase correction for focusing through media inhomogeneities. Some of these methods are based on the theoretical evaluation of sound propagation in a medium. Other methods require application of a hydrophone for signalphase tuning (Clement et al. 2000; Sun and Hynynen 1999) and phase-conjugation (Derode et al. 2002). These methods provide focusing of harmonic signals produced by multi-element arrays. An alternative approach to ultrasound focusing is using the time reversal acoustics (TRA) principles (Fink 1997). It has been shown that TRA ultrasound focusing systems are capable of delivering and steering ultrasound to a chosen location in a heterogeneous medium (Montaldo et al. 2004a). Numerous reflections from boundaries and internal structures, otherwise disturbing conventional geometrical focusing, improve Address correspondence to: Yegor Sinelnikov, ProRhythm Inc., 105 Comac Street, Ronkonkoma, NY 11779. E-mail: yegorasha@ yahoo.com 86

Time reversal acoustic focusing with a catheter balloon d Y. D. SINELNIKOV et al.

This paper presents the results of an experimental investigation of TRA focusing of short pulses using a catheter balloon as the reverberator. Complimentary theoretical modeling is used to compare the performance of short and long signals TRA focusing and recommends a therapeutic catheter design. The main goal of this work is to test the feasibility and quality of TRA focusing using a developed high-intensity focused ultrasound (HIFU) catheter balloon with a single ultrasound transducer. Measuring focal spot dimensions and focal spot steering using the time reversal principle is amongst the main objectives. The HIFU catheter was developed to treat AF (Meininger et al. 2003; Wong et al. 2006) and was found effective in the creation of transmural thermal lesions (Okumara et al. 2008). The approach was based on the demonstrated effectiveness of the pulmonary vein (PV) electrical isolation, which markedly reduces the episodes of AF, common for cardiac arrhythmia (Haissaguerre et al. 1998). The use of HIFU for cardiac ablation (Zimmer et al. 1995; He et al. 1995; Hynynen et al. 1997) and its potential for clinical application (Lee et al. 2000; Engel et al. 2006) prompted the development of the HIFU catheter for AF ablation (Sinelnikov et al. 2009). The HIFU catheter is inserted into a femoral vein and advanced into a left atrium through a trans-septal guiding sheath. First, the sheath is advanced through a vena cava into a right atrium and into the left atrium after transseptal puncture. Second, the HIFU catheter is inserted through the sheath, introduced into the left atrium. Then, the distal catheter balloon is inflated to a working size (Fig. 1). A physician then steers the catheter into position with the targeted PV and performs ablation. The catheter balloon serves as a parabolic reflector that forward focuses ultrasound into a ring of high intensity, which creates a circumferential lesion in PV ostia, with the diameter fully defined by the balloon geometry. Figure 1 illustrates an occlusion of the targeted left superior PV with the HIFU catheter: the interior of the PV is highlighted by an angiographic contrast. In this approach, the size of ostial lesion is fully defined by the geometry of the HIFU catheter balloon. If the PV orifice has a substantially different diameter than that of the balloon, the ablation procedure can be ineffective. An anterior-posterior diameter of the PV ostia varies between 9 and 27 mm according to a computed tomographic depiction study (Jongbloed et al. 2005). This warranted the development of a set of catheters with different size balloons, allowing physicians to make critical decisions about what size catheter to use for a particular PV. As a result, catheter exchanges became a common practice, lengthening the procedure and introducing additional risks to the patient (Reddy et al. 2008). On the basis of clinical experience, the clear and unmet need to develop a method to create circular lesions of different sizes using a single

87

Fig. 1. The angiogram of the HIFU catheter balloon in the human heart. (a) The white dashed arrows outline the ultrasound path from cylindrical transducer inside a water-filled balloon and focused by a parabolic reflector interface (dotted line) into a PV atrial ostia (large arrows). (b) Left superior PV highlighted by angiographic contrast injection. (c) Transseptal delivery sheath. (d) Pacing catheter in coronary sinus vein. (e) Electrical mapping lasso catheter. Image is obtained following an institutional review board approved imaging protocol from the Medizinische Klinik Allgemeines Krankenhaus St. Georg in Hamburg, Germany.

catheter emerged. Moreover, simultaneous isolation of ipsilateral PV and the ability to deliver linear lesions was noted as a critical factor to the success of AF treatment (Satomi et al. 2007). The existing HIFU catheter is unable to produce such accurate tailoring of the acoustic focus to target locations, whereas TRA focusing has the potential to create lesions of multiple size and shape without the need to replace catheters. The time reversal method also optimizes the ultrasound energy deposition at the focus, even if the medium is absorbing (Aubry et al. 2008). In this study, we present the evaluation of using the HIFU catheter balloon as an acoustic reverberator to focus ultrasound energy in TRA mode. As the balloon comes within close proximity of an atrial tissue, the focusing distance of 2 mm has been selected. METHODS The original HIFU catheter balloon was designed to geometrically focus the ultrasound energy redirected by a proximal reflector. To create an acoustic reverberator necessary for accumulating acoustic energy in time and subsequently directing it to the target location using TRA principles, we added an additional distal reflector and a distal transducer as shown in Fig. 2. Two opposite reflectors enhanced reverberation of ultrasound inside the water balloon. A middle section of the water balloon was left transparent to ultrasound. The diameter of the balloon was 24 mm and ~10 mL of liquid was required to inflate it to a predefined shape. When inflated to the level of an operating pressure, the balloon repeated the

88

Ultrasound in Medicine and Biology

Volume 36, Number 1, 2010

Function Generator

Desktop Computer

Power Amplifier

PCI Oscilloscope

Electrical Matching

Reverberator balloon

Hydrophone and x-y scan Fig. 2. Schema of modified HIFU catheter with two transducers. Dashed line marks the location of vertical z-scan with hydrophone.

shape of a mold used to form it by the blow molding process. In our experiments, we used an axially symmetric nylon balloon with smooth surfaces. It was filled with degassed deionized water at typical operating pressure of ~0.3 bars. At 9 MHz, ultrasound attenuation in the balloon walls with a thickness much smaller than the ultrasound wavelength (962 microns vs. 166 microns in water) was negligible. The gas-filled reflector interstices (Fig. 2) provided an acoustic impedance mismatch with the water contained inside the main nylon balloon, which was a precondition for forming a reverberation chamber. Ultrasonic waves were allowed to exit the reverberator though a 5-mm uncovered gap in a middle section of the balloon. We explored two different configurations of the acoustic reverberator: an axially symmetric and an asymmetric configuration. The symmetric reverberator was obtained by inflating water in the main balloon to 0.3 bars, while maintaining reflectors’ interstices at room pressure. At this condition, the walls of the main balloon were smooth. The axially asymmetric reverberator was obtained by reducing the main balloon water pressure and increasing air pressure so that the sections adjacent to the reflectors developed an inhomogeneous wrinkle pattern. The acoustical field mapping experiments were conducted in a water tank using a fine-needle hydrophone. A block diagram of the experimental setup is shown in Fig. 3. The catheter with a dual reflector was submerged into a water tank. Its shaft was held in a rotational stage, allowing the measurement of ultrasound field in the azimuthal direction. The hydrophone was mounted on a translational arm that could be moved vertically and horizontally. TRA focusing of ultrasound was conducted to the location of the hydrophone. The experiments were conducted with both configurations of the reverberator:

Degassed water tank Fig. 3. System setup for performing TRA focusing measurements in water tank.

a fully inflated catheter balloon having smooth axially symmetric reflecting walls, and a slightly inflated balloon, having randomly wrinkled walls that provided additional chaotic scattering of ultrasound waves. Electronic part of an experimental setup was comprised of a signal transmitting electronic system and the 12-bit CompuScope 8249 oscilloscope (DynamicSignals LLC, Lockport, IL, USA) for signal recording. An arbitrary frequency generator 33120A (Agilent Technologies Inc., Santa Clara, CA, USA) was programmed for signal radiation at a sampling frequency of 50 MHz and the same sampling rate was used by the scope. An electrical signal amplified by the Pulsed RF Power Amplifier (Communication Power Corporation, Hauppauge, NY, USA), was applied simultaneously to two cylindrical transducers connected in parallel and located inside the inner balloon. Both transducers had a diameter of ~2.7 mm, length 6 mm and a sharp resonance frequency of ~9 MHz. Transmitted ultrasound signals were measured by a 40-micron needle hydrophone (Precision Acoustics Ltd, Dorchester, UK) and amplified by a hydrophone 25-dB booster. Experiments were performed at relatively low time-averaged electrical power (5 to 10 W at 1 kHz repetition rate) to minimize the effect from possible temperature-induced drift in the electromechanical transmission line and in the piezoelectric transducer characteristics. Standard TRA focusing of ultrasound was achieved in a sequence of three steps. First, a short harmonic pulse with a carrier frequency of 9 MHz and duration of 10 cycles was applied to the

Time reversal acoustic focusing with a catheter balloon d Y. D. SINELNIKOV et al.

transmitting transducers, which radiated corresponding acoustic signals into the balloon reverberator (Fig. 4a, b; grey). Second, a long reverberated signal, resulting from the multiple bouncing of ultrasonic waves within the balloon, was recorded by a hydrophone (Fig. 4a, b; black). Third, the recorded signal was time-reversed using a desktop computer, normalized and programmed back into the frequency generator. The reversed signal was amplified and applied to both transducers simultaneously. The collections of applied (gray) and recorded (black, marked by block arrows) signals are shown in Fig. 4 for a symmetric (left-hand side panels) balloon and an asymmetric with wrinkled walls (right-hand side panels) balloon. The radiated and recorded time-reversed signals are shown in Fig. 4c and 4d. The temporal-peak intensity of the TRA focused signal can be increased by modifying the signal amplitude while preserving the phase (Montaldo et al. 2001). A binary mode with three intensity levels was used in experiments. The signal applied to the transmitters A(t) was converted into a new binary signal Ab(t) according to the following formula: ( Am for A $ kAm Ab 5 0 for 2kAm , A , kAm ; (1) 2 Am for A # 2kAm where Am is the amplitude of the signal and k is the threshold parameter set to 5% in our experiments. The acoustic pressure of the TRA radiated and focused signals in the binary regime is shown in Fig. 4e and 4f. Comparing the values of the attained maximum pressure of the radiated signal shown in the panels of Fig. 4, it is seen that the binary mode leads to a substantial increase by a factor of three or four in the TRA focused signal amplitude. RESULTS TRA focusing over a section of cylindrical surface coaxial with the catheter balloon axis and located at about 2 mm from balloon surface was investigated. The regionof-interest was sampled with a step of about 100 microns using the rotational and translational stages of an automated scanning setup. At each location, the hydrophone signal was recorded over 320 ms using 100 points averaging. A set of signals s(z,q,t) was measured, where t denotes time and z, q are the spatial coordinates. From this set of signals, the temporal-peak acoustic pressure amplitude pðz; qÞ5max{jsðz; q; tÞj} was taken to form the spatial distribution shown in Fig. 5 over a 232 mm ðz; qÞregion. The TRA signal was constructed to focus in the center of the map, which was arbitrarily chosen by manually moving the hydrophone close to the middle of the balloon surface.

89

It was found that for the fully inflated balloon, the axial symmetry of the transducers and balloon reverberator limited TRA focusing to the radial and axial directions and there was no focusing in the azimithal direction (Fig. 5 a, 5c). The TRA system based on a symmetrical reverberator provided a cylindrical focus close to that of the original HIFU catheter. In the case of an asymmetric reverberator, an additional reverberation facilitated by the irregular balloon wrinkles, provided focusing of ultrasound energy in all directions, including azimuthal, and created a spot-like focal pattern (Fig. 5b, 5d) using either standard or binary TRA modes. The quality of the TRA focusing was characterized by calculating the ratio of the spatial-peak temporalpeak intensity ISPTP and the spatial-average temporalpeak ISATP in the region-of-interest. This ratio is defined as the temporal-peak concentration factor mTP : mTP 5

ISPTP : ISATP

(2)

The temporal-peak intensity was calculated from the temporal-peak acoustic pressure as I5p2 =rc, where r is the density and c is the velocity of sound in water. The concentration factor mTP provides a quantitative means of assessing the quality of ultrasound focusing by the TRA system: a higher mTP indicates better quality of the TRA spatial focusing. Table 1 presents the concentration factor for the tested cases evaluated within 131 mm coaxial region at about a 2-mm distance from balloon surface. Based on this rough estimate, the binary mode consistently yields a higher peak-to-background ratio. The concentration factor is the highest for an asymmetric wrinkled balloon, indicating the important role of chaotic scatterers in the TRA focusing. Clearly, the binary regime provided a higher absolute pressure amplitude, relative to a standard mode of TRA focusing, as shown in Fig. 6. At the same time, the binary regime produced a higher spatial concentration factor and a respectively lower level of spatial side lobes. Table 1 summarizes the results in terms of the spatial temporalpeak concentration factor mTP . The gain in the peak acoustic pressure for the binary mode relative to a standard mode was about 200% and 400% in axially symmetric and asymmetric balloon configurations, respectively. One of the advantages of the TRA system is the opportunity to steer the focal spot. For the balloon configuration, we demonstrated this by focusing on various points and measuring the TRA-focused field spatial distribution. The presented TRA tests were conducted at relatively low power, avoiding the heating of transducers and media during long exposure. Implicit to this study is the assumption that TRA focusing is achievable at high power without significant modifications to the transducer and balloon reverberating structure. In our preliminary

90

Ultrasound in Medicine and Biology

Volume 36, Number 1, 2010

Axially symmetric balloon

Randomly wrinkled balloon

b

a 1 .8 k P a

0

100

200 Time, ( s)

300

2 .7 k P a

0

100

200 Time, ( s)

d

c

1 9 .3 k P a

1 2 .8 k P a

0

100

200

300

0

100

Time, ( s)

e

200 Time, ( s)

300

f 2 4 .0 k P a

0

300

100

200 Time, ( s)

300

7 3 .1 k P a

0

100

200 Time, ( s)

300

Fig. 4. TRA signals observed in the experiments. Left panels present signals for the fully inflated balloon and right panels show signals for the case when the inner balloon pressure was reduced to create irregular wrinkles of the reflective interfaces. (a, b) A short pulse (black) applied to the transducers and direct signal (grey) recorded by the hydrophone. (c, d) Time-reversed radiated signals (grey) and the TRA-focused recorded signals (black). (e, f) Time-reversed radiated signals in the binary mode (gray) and the recorded TRA focused signals (black). Numbers in figures show peak-to-peak acoustic pressure maximum amplitudes in received signal.

short-run experiments with high-power TRA, a similar catheter balloon resonator structure was able to achieve the maximal acoustic pressure up to 2 MPa. Mathematical modeling In our experiments, we demonstrated spatial focusing of short ultrasound pulses produced by a narrow-band piezotransducers attached to a water-filled balloon rever-

berator. Meanwhile, therapeutic ultrasound treatments are typically performed not with short pulses, but rather with long continuous signals to deliver a sufficient amount of energy to the target region in the tissue. Because our current experimental setup disallows us to work with long continuous signals, we approached this problem theoretically and developed a mathematical model of the TRA focusing of signals with different duration and frequency content.

Time reversal acoustic focusing with a catheter balloon d Y. D. SINELNIKOV et al.

91

Fig. 5. Two-dimensional acoustic peak intensity maps. Left panels (a, c) present results for fully inflated balloon, and right panels (b, d) for randomly wrinkled balloon. Top panels (a, b) show intensity distribution using standard TRA mode; bottom panels (c, d) show the results of binary TRA mode experiments. The temporal-peak intensity concentration factor m is shown for each panel. The maximum value at the colorbar corresponds to a normalized intensity. Axes are scaled in terms of the wavelength at 9 MHz in water.

The model comprised several point sources located inside a box reverberator with characteristic dimensions a522 mm, b518 mm and c525 mm. A line outside the box constituted a set of target points, where contributions from multiple kaleidoscopic reflections were superimposed, taking into account their time delays. Each reflection signal arriving at the target site was shifted in time proportional to the distance from a respective virtual source mirrored across the reflective resonator’s boundaries. The model is illustrated in Fig. 6a, where for clarity, only one source S is shown with just three kaleidoscopic reflections. To account for attenuation in the medium, the amplitudes of the reflection signals arriving at the target point were taken to be inversely proportional to a respective acoustic path length. The TRA-focused signals in the target points were calculated following several steps. First, a forward signal was calculated by superimposing the signals resulting from the multiple reflections in the reverberator and arriving Table 1. Concentration factor mTP Configuration

Standard

Binary

Axially symmetric Asymmetric wrinkled

2.62 7.47

4.39 11.18

at a target point D. Second, the forward signal at the unique target point D was time reversed, applied to each source and retraced through the resonator to a set of target points. Finally, the maximum intensity of the reversed signals in the target points was evaluated. The spatial intensity distribution was characterized by calculating a temporal-peak intensity ITP ðz; qÞfmax{pðz; q; tÞ2 } and temporal-average intensity ITA ðz; qÞfmean{pðz; q; tÞ2 }, where t denotes time, z and q are the spatial coordinates and pðz; q; tÞ is the received signal at the target points. As a baseline for further evaluation of different waveforms, we characterized the focusing of a short 0.2-ms (two cycles) 9-MHz pulse. The focused time-reversed signal in the target point D is shown in Fig. 6b by a solid black line. Temporal-peak (solid lines) and temporal average (dashed lines) intensity distributions are shown in Fig. 6c. The focal peak dimensions in both cases were close to one wavelength. Yet, as seen in Fig. 6c, the temporal average intensity concentration factor mTA is significantly less than corresponding temporal peak intensity concentration factor mTP . Subsequently, we considered two types of signals: a wide-band signal with frequency bandwidth from 8– 10 MHz (WB) and a narrow-band signal constituting a 9-MHz wave burst (NB). Both signals had the time

92

Ultrasound in Medicine and Biology

Volume 36, Number 1, 2010

Fig. 6. TRA focusing in rectangular resonator: (a) model geometry, and locations of virtual sources for three kaleidoscopic reflections; (b) focused time-reversed signals for a short pulse (black); wide-band long signal (WB: grey) and narrow-band long signal (NB: light grey); (c) spatial distribution of temporal-peak (ITP) and average (ITA) acoustic intensity for a TRA short pulse; (d) spatial distribution of the normalized temporal-average intensity for long WB and NB signals with single and multiple sources.

duration of about 25 ms. The temporal-peak intensity of these signals after time reversal was found to be significantly less than that of a short 0.2-ms pulse. Figure 6b shows that the temporal peak-to-peak pressure of a short pulse is several times higher, whereas the duration of the high-pressure amplitude of a short pulse is significantly shorter than that of the 25-ms long signals. On the other hand, the total energy delivered by long signals is higher than that of a short pulse, which is important for therapeutic applications. The temporal-average intensity distributions were calculated for a single, and several point sources and are shown in Fig. 6d. The single broadband signal (dashed line) produced focusing comparable in quality to a single short pulse (solid line), whereas the single narrow band signal did not focus at all. Increasing the number of sources enabled us to compensate for a deficiency of narrow band time reversal and produce focusing comparable to a short pulse. The dotted line in Fig. 6d shows the temporal-average intensity profile produced by time reversal focusing of the narrow band signal from seven point sources. It compares well with wide band focusing shown by the solid line in Fig. 6d and with temporal-

peak intensity shown by solid line in Fig. 6c. Notably, the temporal-average intensity concentration factor mTA was found to be proportional to the number of sources and independent of the signals’ frequency bandwidth. Indeed, in the focus, the wave constructive interference results in a signal amplitude proportional to the number of sources and intensity proportional to the square of the number of sources. Outside the focus, different contributions are not correlated and intensity becomes a linear function with respect to the number of sources. This important result of the modeling suggests that TRA focusing can be achieved in a continuous regime using multiple sources of limited bandwidth, a feasible option in a catheter design that would require a sectioning of transducers and multiple electrical leads. DISCUSSION We demonstrated experimentally the feasibility of TRA focusing of ultrasonic waves in a pulsed regime using liquid-filled balloon reverberator. A narrow band 9.060.25-MHz transducer delivering short, 2– 3 cycles of ultrasound, pulses were used. Time reversal

Time reversal acoustic focusing with a catheter balloon d Y. D. SINELNIKOV et al.

experiments were performed using single channel and the TRA focusing was evaluated for two different configurations of the balloon resonator. We observed that the irregular wrinkles of the balloon resonator walls improved TRA focusing of ultrasound. The dimension of the wrinkles was comparable or smaller than the wavelength of ultrasound at 9 MHz, working as effective scatterers. The single-channel ultrasound catheter with smooth surface axially symmetric balloon reverberator was capable of creating axially symmetric ultrasound ring patterns in the TRA regime. An asymmetrically wrinkled balloon produced better ultrasound focusing, and made possible to overcome the constraints of balloon axial symmetry to produce localized focal spots instead of rings. The latter was possible because of the presence of small-scale wrinkles that scattered and mixed ultrasound waves from cylindrical sources. Although the wrinkle pattern of the balloon walls was not reproducible in the present experiment, we hypothesize that the balloon wall irregularities could be formed by introducing a random groove pattern in a surface of a metal balloonforming mold. During the blow-molding process, polymer film would reproduce the solid surface pattern, which would retain shape when such a balloon is inflated, providing a means of constructing asymmetric balloons. Because the evaluation of focusing of the long signals was not feasible in the current experimental setup, we conducted complimentary mathematical modeling of TRA focusing and determined that for long signals the spatialpeak temporal-average intensity ISPTA and respective temporal-average concentration factor mTA 5ISPTA =ISATA can be substantially increased by increasing the number of transmitting sources. The intensity ISPTA was dependent on the resonator geometry, signal shape and bandwidth. On the other hand, the concentration factor mTA was found to be linear function of the number of sources and independent of the signal shape and bandwidth, suggesting feasibility of the TRA focusing using long signals and multiple transducers. This suggests that a multichannel catheter can be developed to increase the overall system power output and improve the temporal-average intensity concentration factor of the TRA ultrasound focusing. A disadvantage of the current TRA system is that it requires a hydrophone (‘‘a beacon’’) for focusing. Obviously, the hydrophone cannot be used in a human heart, making it challenging to register intended lesions relative to the catheter balloon. There are several possible solutions to this problem. First, it is possible to calibrate the TRA balloon-focusing system in a water tank and create a library of signals providing ultrasound focusing to the predefined locations around the catheter. Taking into account the relatively short propagation path between balloon and a tissue in AF isolation, we anticipate that the same library of signals will focus ultrasound energy

93

through blood in atria. At a depth of a few millimeters, ultrasound absorption can be compensated by transmitting the signal in binary mode and it is plausible that signals, saved in a computer, will still focus at a target location in the myocardial wall. Thus, the TRA technique will enable physicians to focus ultrasound accurately tailored to the patient’s anatomy. The orientation of the catheter balloon with respect to the atrial anatomy can be guided by fluoroscopically visible landscape markers installed on the surface of the balloon reverberator. The emitting transducers can also record reverberated signals and match them against a pre-recorded library of signals to ensure consistent and repeatable TRA focusing and avoid ultrasound exposure of the surrounding tissues. Testing the feasibility of these approaches and exploring other alternatives will be a long-term goal of further research. CONCLUSIONS The conducted experiments demonstrated efficient TRA focusing of ultrasound in both time and space using a modified catheter balloon. Hydrophone measurements showed that the dimension of the TRA focal spot approaches the diffraction limit. Among the two studied configurations, axially symmetric and asymmetrical, results demonstrate that balloon reverberator surface irregularities enhance TRA focusing of ultrasound. Furthermore, experiments demonstrated that TRA focusing could be improved using binary signals producing higher spatial-peak temporal-average ultrasound intensity ISPTA and yielding a higher concentration factor mTA than a standard TRA mode. Another possibility for increasing the energy output of the TRA focusing of ultrasound by the balloon reverberator follows from the results of mathematical modeling, which showed that ISPTA is proportional to the number of sources squared. Increasing the number of sources can be achieved by adding new or sectioning existing transducers and adding time reversal channels. Acknowledgment—The authors would like to thank Leslie Le-Passoff for technical assistance and Reinhard Warnking for management support. This research was funded by Prorhythm, Inc.

REFERENCES Anderson BE, Griffa M, Larmat C, Ulrich TJ, Johnson PA. Time reversal. Acoust Today 2008;4:5–11. Aubry JF, Pernot M, Marquet F, Tanter M, Fink M. Transcostal highintensity-focused ultrasound: Ex vivo adaptive focusing feasibility study. Phys Med Biol 2008;53:2937–2951. Clement GT, White J, Hynynen K. Investigation of a large-area phased array for focused ultrasound surgery through the skull. Phy Med Biol 2000;45:1071–1083. Derode A, Tourin A, Fink M. Time reversal versus phase conjugation in a multiple scattering environment. Ultrasonics 2002;40:275–280. Haı¨ssaguerre M, Jaı¨s P, Shah DC, Takahashi A, Hocini M, Quiniou G, et al. Spontaneous initiation of atrial fibrillation by ectopic beats originating in the pulmonary veins. N Engl J Med 1998;339:659–666.

94

Ultrasound in Medicine and Biology

He DS, Zimmer JE, Hynynen K, Marcus FI, Caruso AC, Lampe LF, et al. Application of ultrasound energy for intracardiac ablation of arrhythmias. Eur Heart J 1995;16:961–966. Hynynen K, Dennie J, Zimmer JE, Simmons WN, He DS, Marcus FI, et al. Cylindrical ultrasonic transducers for cardiac catheter ablation. IEEE Trans Biomed Eng 1997;44:144–151. Engel DJ, Muratore R, Hirata K, Otsuka R, Fujikura K, Sugioka K, et al. Myocardial lesion formation using high-intensity focused ultrasound. J Am Soc Echocardiogr 2006;19:932–937. Fillinger L, Sutin A, Sarvazyan A. Time reversal focusing of short pulses. IEEE Ultrasonics Symposium N Y 2007;220–223. Fink M. Time-reversal acoustics. J Phys 2008;118:1–28. Fink M. Time reversed acoustics. Phys Today 1997;50:34–39. Fink M, Montaldo G, Tanter M. Time-reversal acoustics in biomedical engineering. Annu Rev Biomed Eng 2003;5:465–497. Jongbloed MR, Dirksen MS, Bax JJ, Boersma E, Geleijns K, Lamb HJ, et al. Atrial fibrillation: Multi-detector row CT of pulmonary vein anatomy prior to radiofrequency catheter ablation—Initial experience. Radiology 2005;234:702–709. Lee LA, Simon C, Bove EL, Mosca RS, Ebbini ES, Abrams GD, et al. High intensity focused ultrasound effect on cardiac tissues: potential for clinical application. Echocardiography 2000;17:563–566. Meininger GR, Calkins H, Lickfett L, Lopath P, Fjield T, Pacheco R, et al. Initial experience with a novel focused ultrasound ablation system for ring ablation outside the pulmonary vein. J Interv Card Electrophysiol 2003;8:141–148. Montaldo GR, Tanter M, Fink M. Real time inverse filter focusing through iterative time reversal. J Acoust Soc Am 2004a;115: 768–775. Montaldo G, Palacio D, Tanter M, Fink M. Time reversal kaleidoscope: A smart transducer for three dimensional ultrasonic imaging. Appl Phys Lett 2004b;84:3879–3881. Montaldo G, Roux P, Derode A, Negreira C, Fink M. Generation of very high pressure pulses with 1-bit time reversal in a solid waveguide. J Acoust Soc Am 2001;110:2849–2857. Okumura Y, Kolasa MW, Johnson SB, Bunch TJ, Henz BD, O’Brien CJ, et al. Mechanism of tissue heating during high intensity focused ultrasound pulmonary vein isolation: Implications for atrial fibrillation ablation efficacy and phrenic nerve protection. J Cardiovasc Electrophysiol 2008;19:945–951.

Volume 36, Number 1, 2010 Quieffin N, Catheline S, Ing RK, Fink M. Real-time focusing using an ultrasonic one channel time-reversal mirror coupled to a solid cavity. J Acoust Soc Am 2004;115:1955–1960. Reddy VY, Neuzil P, D’Avila A, Laragy M, Malchano ZJ, Kralovec S, et al. Balloon catheter ablation to treat paroxysmal atrial fibrillation: What is the level of pulmonary vein isolation. Heart Rhyth 2008;5: 353–360. Satomi K, Ouyang F, Kuck KH. How to determine and assess endpoints for left atrial ablation. Heart Rhyth 2007;4:374–380. Sinelnikov Y, Fjield T, Sapozhnikov OA. The mechanism of lesion formation by ultrasound ablation catheter for treatment of atrial fibrillation. Acoust Phys 2009;55:1–12. Sinelnikov Y, Sutin A, Zou Y, Sarvazyan A. Time reversal acoustic focusing with liquid resonator for medical applications. Transaction of 6th International Symposium on Therapeutic Ultrasound International Society for Therapeutic Ultrasound, Oxford, UK, August 30– September 2, 2006:82–86. Sun J, Hynynen K. The potential of transskull ultrasound therapy and surgery using the maximal available surface area. J Acoust Soc Am 1999;105:2519–2527. Sutin A, Libbey B, Kurtenoks V, Fenneman D, Sarvazyan A. In: Broach JT, Harmon RS, Holloway JH Jr., (eds). Detection and Remediation Technologies for Mines and Minelike Targets XI. Proc SPIE; 2006. p. 398–409. Sutin A, Johnson P. In: Thompson DO, Chimenti DE, (eds). Review of Quantitative Nondestructive Evaluation. New York: AIP; 2005. p. 385–392. Sutin A, Sarvazyan A. Spatial and temporal concentrating of ultrasound energy in complex systems by single transmitter using time reversal principles. In Proceedings of World Congress on Ultrasonics September 7-10, 2003, Paris, France; pp. 863–866. Tanter M, Pernot M, Aubry JF, Fink M. Compensating for bone interfaces and respiratory motion in high-intensity focused ultrasound. Intern J Hypertherm 2007;23:413–415. Wong T, Markides V, Peters N, Davies D. Anatomic left atrial circumferential ablation to electrically isolate pulmonary veins using a novel focused ultrasound balloon catheter. Heart Rhyth 2006;3:370–371. Zimmer JE, Hynynen K, He S, Marcus F. The feasibility of using ultrasound for cardiac ablation. IEEE Trans Biomed Eng 1995;42: 891–897.