† Corresponding author. E-mail:
The influence of the inhomogeneous tissue layer on the generation of acoustic vortices (AV) is studied theoretically and experimentally based on the phase screen model. By considering the time-shift of a random phase screen, the formula of acoustic pressure for the AV beam generated by a circular array of eight planar piston sources is derived. With the actual correlation length of the abdominal wall, numerical simulations before and after the insertion of the inhomogeneous tissue layer are conducted, and also demonstrated by experimental measurements. It is proved that, when the thickness variation of the phase screen is less than one wavelength, no significant influence on the generation of AVs can be produced. The variations of vortex nodes and antinodes in terms of the location, shape, and size of AVs are not obvious. Although the circular pressure distribution might be deformed by the phase interference with a larger thickness variation, AVs can still be generated around the center axis with perfect phase spirals in a reduced effective radius. The favorable results provide the feasibility of AV generation inside the human body and suggest the application potential of AVs in object manipulation for biomedical engineering.
After the concept of wave front dislocation[1] was introduced by Nye in 1974, the physical meaning of the phase dislocation or the phase singularity in wave propagation attracts more and more attention to the optical vortex (OA). Due to the characteristics of the polarization and the rotational angular momentum of light, the unique structure of phase in OAs was demonstrated to possess the orbital angular momentum (OAM), which could be used to manipulate objects with the exerted rotation torque. In 1992, driven by the OAM transferred from a Laguerre–Gaussian laser beam,[2] the particles were observed to rotate around the beam axis successfully. Through the experimental measurement of the torque, Allen also proved that the OAM of an OA was proportional to the corresponding topological charge. Then, the basic theory of OA was explained in detail by Neuma[3] and the structure design of optical tweezers was also proposed to improve the effects in optical manipulation.
Compared to the light beam, acoustic waves can propagate into deep media and can be used to manipulate particles inside tissues, which suggest the feasibility of broad applications in nondestructive testing and biomedical engineering. By using two sets of phased ultrasonic transducer arrays, the accurate position control of standing wave node[4] in air was realized by Hoshi and the practical application of three-dimensional (3-D) manipulation for mm-scale particles was also demonstrated. Then, based on the surface resonance of phononic crystals, an acoustic sieve[5] was designed by Li to locate, capture, sift and transport nanoscale particles with a local high radiation force. However, the application of the proposed methods is still hindered by the specific manner of the acoustic field.
As a kind of acoustic wave with continuous helical wave fronts around the beam axis, the acoustic vortex (AV)[6–8] shows an obvious pressure zero at the vortex center with an intrinsic phase singularity. The possibility of the OAM transfer to objects is produced by the continuous phase spiral. The object manipulation inside tissues can be realized by the exerted radiation torque, which shows significant application potential in biomedical and industrial areas. An underwater ultrasonic calibration system[9] composed of four high-frequency transducers was designed. The existence of screw wave fronts and phase dislocations was proved and the qualitative relationship between the acoustic pressure and the OAM of the AV was also obtained. A quantitative test[10] for the OAM transfer from the AV to an acoustic absorbing object immersed in a viscous liquid was conducted by Anhauser. In addition, Demore[11] proved that the ratio of the OAM to the acoustic power of an AV was equivalent to the ratio of the topological charge l to the angular frequency ω.
In order to generate AVs with sparse sources, a phase-coded approach[12] was proposed by Yang using a circular array of point sources. It was proved that, for N point sources, the AV beam with a controllable topological charge can be generated, and the phase difference resolution of the adjacent sources was demonstrated to be π. The maximum topological charge of the AV beam generated by an N-element system was proved to be fix[(
In previous studies, the generation of AVs was investigated only in a single medium without the consideration of acoustic transmission in layered or inhomogeneous media. However, the acoustic propagation can be influenced by the inhomogeneity of tissues, resulting in great impact on AVs. In biomedical applications, due to the layered structure and the inhomogeneity of tissues, the abdominal wall should not be treated as a single homogeneous medium. In the field of high-intensity focused ultrasound (HIFU) therapy, the layered model[16–18] of tissues was established to study the influence of layered media on the pressure distribution in the focal area; whereas, the structure of the model is different from that of the abdominal wall, and the phase distortion in acoustic propagation is sensitive[19,20] to the phased array in medical devices. It was reported by Liu that the inhomogeneity of the abdominal wall has obvious influences on the location, shape, and intensity of the focal region, and the impact of the phase distortion introduced by the time delay is more significant[21–23] than that caused by acoustic attenuation in tissues. For the phase-coded approach, besides the physical characteristics of the circular transducer array, the generation of AVs is determined by the accurate individual phase control for each source. Since the acoustic pressure at each point is the superimposition of the signals transmitted from the sources, the distribution of the AV will be affected by the phase shift during the acoustic transmission. Obvious location deviation and pressure deformation of AVs will be produced by the range of the random phase variation. Therefore, the influence of the inhomogeneous tissue layer should be taken into account on the generation of AVs, especially for the abdominal wall in human body.
In this paper, based on the phase variation of the abdominal wall, the influence of the inhomogeneous tissue layer on the generation of AVs is investigated using the random phase screen model. For the AV beam generated by a circular array of planar transducers, the inhomogeneous tissue layer with the actual correlation length of the abdominal wall is applied to introduce the additional random phase shift. The simulation results are also demonstrated by the experimental measurements with an eight-element transducer array and a phase screen slab. The good agreement between numerical simulations and experimental measurements prove that the phase screen has a more obvious impact on the distribution of acoustic pressure than that of phase. When the thickness variation is less than one wavelength, no significant influence on the generation of AVs can be produced. The variations of vortex nodes and antinodes in terms of the location, shape, and size of AVs are not obvious. Although the circular pressure distribution might be deformed by the thickness variation larger than one wavelength, AVs can still be generated around the beam axis with perfect phase spirals. The favorable results certify the feasibility of AV generation inside the human body and suggest the application potential of multiple traps in object manipulation.
To generate the AV beam at the frequency of MHz in water, the phase-coded approach is employed to excite the N-element circular transducer array, which is illustrated in Fig.
By setting the center of the source plane as the origin in the cylindrical coordinates as shown in Fig.
It can be observed that the radiation pattern of each source is determined by ka. For
Although the multiple traps of the AV beam along the center axis can be generated by the main lobe and the side lobes in water, they are also influenced by the complex structure of the abdominal wall with obvious inhomogeneity. Through one-dimensional (1-D) fluctuation measurement of time delay for the abdominal wall, Krammer[25] observed that the arrival time of the wave front was different from each other with random individual difference. The time delay followed the statistic law[26] with a correlation length of 2–10 mm. Hinkelman[27] demonstrated that the time-shift aberration formed by the abdominal wall could be explained by the simplified phase screen model,[28] and could be estimated by a time-shift compensation algorithm. The inhomogeneous tissue layer can be regarded as a thin layer for acoustic transmission, and the random phase fluctuation of acoustic wave is only affected by the influence of the acoustic velocity, while the pressure attenuation is considered as a constant. Based on the correlation length and the mean square deviation (standard deviation) parameter,[21,22] the phase screen is generated randomly by a two-dimensional (2-D) phase interference data.
By considering the inhomogeneity of a tissue layer, the phase screen is applied to calculate the distribution of acoustic pressure. Assume that the attenuation coefficient and the acoustic speed in tissues are higher than those in water with the differences of
Hence, the transmission distance
According to the parameters of the experimental setup, numerical studies of acoustic pressure and phase of the AV beam with a = 0.7 cm, R = 3 cm, and N = 8 were conducted at f = 1 MHz. The phase difference of the adjacent sources was set to
The axial pressure profile of the AV beam generated by the transducer array from z = 10 cm to 40 cm is illustrated in Fig.
In order to analyze the characteristics of the vortex antinodes, the cross-sectional distributions of pressure and phase at z = 40 cm are plotted in Figs.
It was also reported by Hinkelman[27] that the body wall played the most important role among all the phase fluctuations introduced by the inhomogeneity inside the human body. The spatial distribution of the phase screen was determined by the correlation length, which was the inhomogeneous length scale of tissues. Liu[22] also observed that the inhomogeneity of human tissues had a certain impact on the distribution of a focused ultrasound, while the influence of the correlation length was relatively small. Therefore, in this study, an inhomogeneous tissue layer with a correlation length of 3.0 mm and a controllable thickness variation was used to investigate the influence on the generation of AVs. In numerical studies, a phase screen model is used to depict the extended random medium with a series of slabs at a random thickness. The phase variation can be calculated by the fluctuation of the random thickness introduced time-shift. When the thickness is the order of the correlation length of the inhomogeneous tissue layer, the random phase variation is produced by the projection onto a screen without the consideration of acoustic attenuation, which exhibits a Gaussian correlation function with a zero mean. The simulation model of the phase screen with the maximum thickness variation of one wavelength (
By considering the time-shift of the thickness variation, the pressure distribution was simulated with the insertion of the inhomogeneous tissue layer based on Eq. (
For the vortex antinode at z = 40 cm in the transverse plane, the acoustic pressure at the origin is almost zero as shown in Fig.
According to the parameters of the simulation model in Fig.
With the insertion of the tissue layer, the experimental axial pressure profile illustrated in Fig.
The phase screen model used in this study is an efficient approximation method to deal with the influence of random media on wave propagation, providing a clear physical concept for phase interference. The model separates the influences of amplitude attenuation and phase shift from each other to realize theoretical simplification. For the acoustic transmission through an actual inhomogeneous tissue layer, the impact of the random phase screen on the acoustic pressure and phase can be produced by the difference of the thickness. With the random phase screen, the variation of acoustic attenuation is not obvious, while the phase change caused by the acoustic speed is relatively significant with an intensive phase fluctuation. The generation of AVs with the circular transducer array is the superimposition result of the phased acoustic beams, which is more sensitive to the phase variations of the radiated waves. Thus, the random phase screen model is sufficient to analyze the influence of the tissue layer on the generation of AVs inside human body. However, by considering the effects of the acoustic attenuation and the phase shift of the abdominal wall, more accurate study should be further carried out, which shows its great significance to the practical application of AVs in biomedical engineering.
Also as we know, the cross section of an ideal AV has a circular pressure distribution and a helical phase distribution. The time-averaged flux of the OAM density around each circumference in the transverse plane is the same with a same circular distribution of the acoustic radiation force, which is beneficial for the accurate design of particle manipulation. Although the ideal circular pressure distribution might be deformed by the influence of the phase screen with random phase shifts, the formation of AVs can also be proved by the helical phase distributions. The effects of object manipulation and particle accumulation can still be realized by the helical direction of the acoustic radiation forces around each circumference, in spite of the difference of OAMs. Moreover, in order to reduce the influence of the inhomogeneous abdominal wall, a lower frequency of acoustic signals with a bigger wavelength should be applied to achieve the relatively smaller phase differences of the random phase screen. The deformation of AVs introduced by the phase screen can also be reduced by the cancellation effect of phase interference for more acoustic sources of the circular array.
The experimental measurements exhibit good agreements with the numerical results, demonstrating the formation of AVs along the beam axis before and after the insertion of the inhomogeneous tissue layer. However, obvious differences can still be observed between the experimental and theoretical results, which might be introduced by the uncertain factors as below. Firstly, the error is produced probably by the asymmetric structure of the transducer array and the inconsistency of the sources. Secondly, since the results are influenced by the randomness of the inhomogeneous tissue layer, the acoustic distribution of each experimental measurement is unique, which is different from the others. In addition, further investigations on the correlation length impact of the tissue layer and the optimization algorithm[29] of the random phase screen should also be performed for AVs in the practical application of biological tissues.
In this paper, based on the acoustic radiation of a circular array of planar piston sources, the mechanism underlying the generation of AVs is investigated with the consideration of the random phase screen. By applying the actual parameters of the abdominal wall, numerical simulations before and after the insertion of the inhomogeneous tissue layer are performed, and also demonstrated by experimental measurements. Both theoretical and experimental results prove that the distribution of AVs can be affected by the phase screen with random time-shifts. In the actual thickness variation range of human tissues, no obvious influence of the phase screen can be introduced to the generation of AVs, and the variations of vortex nodes and antinodes in terms of the location, shape, and size are not significant. Although the circular pressure distribution might be deformed by the phase interference with the thickness variation higher than one wavelength, AVs can still be generated with perfect phase spirals around the center axis with a reduced effective radius. The favorable results provide the feasibility of the generation of AVs through the abdominal wall, and suggest the application potential of object manipulation inside the human body in biomedical engineering.
[1] | |
[2] | |
[3] | |
[4] | |
[5] | |
[6] | |
[7] | |
[8] | |
[9] | |
[10] | |
[11] | |
[12] | |
[13] | |
[14] | |
[15] | |
[16] | |
[17] | |
[18] | |
[19] | |
[20] | |
[21] | |
[22] | |
[23] | |
[24] | |
[25] | |
[26] | |
[27] | |
[28] | |
[29] |