Acoustic streaming: a new technique for assessing adnexal cysts

Ultrasound Obstet Gynecol 2003; 22: 74 78 Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/uog.156 Acoustic streaming: a new technique for assessing adnexal cysts A. EDWARDS, L. CLARKE, S. PIESSENS, E. GRAHAM and P. SHEKLETON Ultrasound Department, Monash Medical Centre, Southern Health, Clayton Road, Clayton, Victoria, Australia KEYWORDS: acoustic streaming; adnexal cysts; endometrioma; ovarian cysts ABSTRACT Objectives To determine whether acoustic streaming has clinical value in the differentiation between various ovarian and adnexal cysts. Methods We assessed 29 adnexal cysts, for which pathological diagnosis was available, for the presence of acoustic streaming during B-mode and color sonographic evaluation. Results Acoustic streaming was detected in 15 (52%) of the cysts. The most common cyst, endometrioma (n = 7), did not exhibit acoustic streaming in any case, while of the remaining 22 cysts, 15 exhibited acoustic streaming (P = 0.0017). Dermoid cysts exhibited acoustic streaming in two of six (33%) cases. In addition acoustic streaming was noted in two of two (100%) hemorrhagic cysts, eight of ten (80%) cystadenomas, two of three (67%) malignant cysts and in the one abscess. Conclusions Acoustic streaming is the first sonographic feature that may be able to completely exclude endometrioma as a possible diagnosis for an adnexal cyst. Copyright 2003 ISUOG. Published by John Wiley & Sons, Ltd. INTRODUCTION Assessment of adnexal cysts is a major indication for gynecological ultrasound. The ability to accurately differentiate not only between benign and malignant lesions, but also between different types of benign cysts, allows optimization of management whether it is expectant or surgical in nature. Decisions regarding the most appropriate pelvic surgical technique, such as laparoscopy vs. laparotomy, can be greatly affected by preoperative information regarding the likelihood of malignancy, or knowledge about which type of benign cyst is most likely to be present. For example, the two most common types of persistent ovarian cyst in premenopausal women, endometriomas and dermoid cysts, can be difficult to assess using standard techniques. The presence of solid components within dermoid cysts means that they will often be scored as highly suspicious of malignancy by morphological scoring systems 1, while the wide range of sonographic appearances of endometriomas makes their morphological assessment complex and confusing, with a number of authors publishing different ultrasound criteria for their diagnosis 2. This study explores a new sonographic technique, acoustic streaming, for assessing ovarian cysts. Acoustic streaming describes the movement of fluid or air due to a sound field. It is caused by a transfer of energy from sound waves, and results in movement in the direction of wave propagation 3,4. The transfer of energy occurs as the sound wave strikes reflecting and absorbing obstacles in its path and hence is related to attenuation. It was initially termed quartz wind by Meissner in 1926, when it was noted that ultrasonic beams caused air movement 5, but recently the movement of fluid in biological situations has received more attention. While acoustic streaming has previously been noted in cyst fluid, in the lateral ventricles of a premature infant, and in blood 5, there has been very little evaluation of its clinical role. Nightingale et al. 6,7 demonstrated that acoustic streaming is able to help differentiate cystic from solid breast lesions, concluding that the method would be a useful adjunct to conventional sonography, while Shi et al. 8 used a hematoma model to demonstrate that acoustic streaming detection may be valuable for distinction between liquid and clotted blood. We decided to investigate whether acoustic streaming could help in the diagnosis of ovarian and adnexal cysts. To our knowledge this is the first study to address this issue. Specifically, our aims in this pilot study were primarily to determine whether different ovarian and adnexal cysts exhibit different acoustic streaming characteristics, Correspondence to: Dr A. Edwards, 14-16 Dixon Street, Clayton 3168, Victoria, Australia (e-mail: andrew1000@bigpond.com) Accepted: 19 February 2003 Copyright 2003 ISUOG. Published by John Wiley & Sons, Ltd. NEW TECHNIQUES

Acoustic streaming 75 and secondarily to evaluate whether any differences in acoustic streaming characteristics may have a clinical value. METHODS As this was a pilot study with no previous data available we initially set a period of 6 months to collect data, estimating that we would be able to study at least 40 ovarian or adnexal cysts in women presenting to our ultrasound department. During this time, upon detection of an adnexal or ovarian cyst during routine pelvic scanning, we prospectively analyzed the cyst for the presence of acoustic streaming according to the following protocol. We used an ATL HDI 5000 ultrasound machine (Philips Medical Systems, Best, The Netherlands) with a C8-4v transvaginal probe. Settings were the standard factory default for pelvic scanning. All women with an ovarian or adnexal cyst that contained some internal echoes on B-mode scanning, and that was greater than 2.5 cm in diameter, were consecutively included. Women with totally solid lesions or cysts with anechoic contents were excluded. All women were initially examined transabdominally and transvaginally following our department s protocol for a pelvic ultrasound examination. When a cyst meeting the above criteria was detected, it was assessed transvaginally for the presence of acoustic streaming (Figure 1). The transducer was held still until any movement within the cyst had ceased. The color box was then applied to all or a part of the cyst with the color gain low enough to enable a clear gray-scale view of the cyst contents. We classified acoustic streaming as being present if movement of particles within the cyst away from the transducer was visible on the grayscale image. To double-check we then turned off the color box and waited for movement to cease before reapplying it to check for streaming once more. For the purposes of this study we then subjectively classified the acoustic streaming velocity to be slow, medium, or fast. For each cyst we also checked for streaming in all areas of the cyst. For each cyst we then recorded the standard sonographic description along with the sonographic diagnosis if one was made. For patients who underwent surgery for examination, biopsy or removal of the cyst, operative notes and pathology reports were used to verify the diagnosis. Cysts for which histological analysis was not performed, such as when surgical excision was not considered necessary, were excluded from further evaluation within the study. In patients with multiple or complex adnexal masses, care was taken to correlate sonographic and pathological findings. Statistical analysis involved comparison of proportions using Stata 6.0 (Stata Corporation, College Station, TX, USA). RESULTS Fifty-six adnexal cysts in 52 women met the inclusion criteria. At the time of writing 29 of the cysts had been Particles Transducer Beam Cyst Figure 1 Schematic diagram showing principles of assessment of acoustic streaming; acoustic streaming was recorded as present when the cyst particles were seen to move directly away from the transducer when the color box was applied. managed surgically allowing pathological diagnoses. The remaining 27 cysts were excluded from further evaluation. Comparison between ultrasound and pathological diagnoses is made in Table 1. The breakdown of the presence of acoustic streaming according to histological diagnosis is shown in Table 1 and is summarized in Table 2. Acoustic streaming was detected in 15 (52%) of the 29 cysts (Table 2). In 12 of these cases, the velocity of acoustic streaming was subjectively classified as medium. The most common cyst, endometrioma (n = 7), did not exhibit acoustic streaming in any case (Figure 2). One cyst with the sonographic features of a dermoid cyst was noted to exhibit acoustic streaming only in a confined area. Histopathological correlation showed hemorrhage in this area and hence this cyst was included as a hemorrhagic cyst for the purposes of analysis in this study. There were six dermoid cysts and two (33%) of these displayed acoustic streaming, although in one the streaming was

76 Edwards et al. Table 1 Details of each adnexal cyst Size (cm) Sonographic diagnosis Pathological diagnosis Streaming 13 Dermoid Hemorrhage within dermoid Yes 7 Dermoid Dermoid No 5 Hemorrhagic Hemorrhagic Yes 9 Mucinous cystadenoma Mucinous cystadenoma Yes 9 Endometrioma Endometrioma No 3 Endometrioma Endometrioma No 3 Endometrioma Endometrioma No 5 Endometrioma Endometrioma No 14 Adenocarcinoma Endometrioid adenocarcinoma Yes 5 Endometrioma Endometrioma No 3 Dermoid Dermoid No 4 Dermoid Dermoid No 8 Cystadenoma Mucinous cystadenoma Yes 11 Cystadenoma/carcinoma Dermoid Yes 7 Dermoid Dermoid No 9 Endometrioma Endometrioma No 14 Cystadenoma/carcinoma Mucinous adenocarcinoma Yes 4 Cystadenocarcinoma Cystadenoma (+ Brenner) No 4 Dermoid Dermoid Yes 7 Endometrioma Endometrioma No 6 Cystadenoma Mucinous cystadenoma Yes 9 Cystadenoma Chronic tubo-ovarian abscess Yes 9 Cystadenoma Cystadenofibroma Yes 9 Cystadenoma Mucinous cystadenoma Yes 20 Cystadenocarcinoma Clear cell tumor (malignant) No 5 Cystadenoma Serous cystadenoma Yes 8 Mucinous cystadenoma Mucinous cystadenoma Yes 5 Serous cystadenoma Serous cystadenoma Yes 9 Mucinous cystadenoma Mucinous cystadenoma No Table 2 Summary of presence of acoustic streaming according to diagnosis Cyst type n Acoustic streaming (n (%)) Endometrioma 7 0 (0) Dermoid 6 2 (33) Cystadenoma 10 8 (80) Hemorrhagic 2 2 (100) Malignant 3 2 (67) Abscess 1 1 (100) Total 29 15 (52) subjectively very slow, moving only about 1 cm in 30 s. Two cystadenomas did not display acoustic streaming: one had a significant solid component with Brenner cell features while the other was a mucinous cystadenoma (Figure 3). In addition, two cystadenomas displayed slow acoustic streaming (a cystadenofibroma and a mucinous cystadenoma). Both serous cystadenomas in the study displayed acoustic streaming (Figure 4). DISCUSSION The primary aim of this pilot study was to determine whether different types of ovarian and adnexal cysts have differing characteristics with respect to acoustic streaming. We have clearly shown that this is the case. In our study no Figure 2 Ultrasound image of an endometrioma. With the color box applied, the cyst particles are observed in B-mode to demonstrate movement away from the transducer. In all cases of endometrioma, acoustic streaming was absent. endometriomas exhibited acoustic streaming, while 15 of the remaining 22 cysts (67%) displayed acoustic streaming (P = 0.0017). Dermoid cysts exhibited acoustic streaming in two (33%) cases. In addition acoustic streaming was noted in two (100%) hemorrhagic cysts, eight (80%) cystadenomas, and two (67%) malignant cysts. A secondary aim was to evaluate whether any differences in acoustic streaming characteristics may have a clinical value. Endometriomas have been well

Acoustic streaming 77 Figure 3 Ultrasound image of a mucinous cystadenoma. In this multiloculated cyst, acoustic streaming was absent in all areas. This was the only mucinous cystadenoma not to display acoustic streaming. Figure 4 Ultrasound image of a serous cystadenoma. In real-time the particles within this cyst were seen to move steadily away from the transducer when the color box was applied but their movement quickly ceased as soon as the color box was moved away. documented to have a broad range of sonographic appearances 2,9, which can make it difficult to distinguish them from other adnexal masses. Various authors have established sonographic criteria for identifying endometriomas 10 13, claiming positive likelihood ratios from 8 to 37 and negative likelihood ratios from 0.1 to 0.2. Acoustic streaming is, to our knowledge, the first sonographic feature that may allow complete exclusion of endometrioma in the diagnosis of adnexal cyst. Furthermore, although we have not yet formally assessed intra- and interobserver error, the objective assessment that acoustic streaming is either present or absent is generally straightforward, making the technique a potentially valuable addition in the assessment of adnexal masses. At this stage we have not been able to explore whether some subtle features of acoustic streaming may provide further diagnostic value in the assessment of adnexal cysts. While subjective assessment of acoustic streaming velocity did not provide additional useful information in this study, it has previously been established that acoustic streaming velocity is related to the soundwave pulse intensity, fluid attenuation and fluid viscosity 6,14. The velocity is directly or closely proportional to the pulse intensity 5,8,15.Inthis study we used the standard default settings for pelvic scanning as set by the manufacturer (ATL). According to the acoustic output table produced by ATL 16, the maximum derated pulse average intensity at these settings is 252 mw/cm 2, corresponding to a mechanical index (MI) of 1.24. It has been suggested that for blood a threshold exists below which streaming is not seen 8 ;this may well also apply to other fluids such as that in adnexal cysts. We have initiated further work to explore the effect of output intensity on both the presence and the velocity of acoustic streaming in adnexal cysts. Acoustic streaming is directly proportional to fluid attenuation 14, presumably due to attenuation caused particularly by the absorption and reflection of particles that the sound wave strikes and to which it transfers momentum as it travels through the fluid 3. In this study we included only cysts that contain particles visible in B-mode ultrasound. Generally we have found the relationship of acoustic streaming to attenuation to be the opposite of that described above: we found that cysts with relatively few low-level echoes, such as cystadenomas, tended to display free flowing streaming, while cysts filled with echoes, such as endometriomas, displayed no streaming. This appears to highlight fluid viscosity as an overriding factor. Streaming velocity is inversely proportional to fluid viscosity 14. We are not aware of any data describing the relative viscosities of fluids within adnexal cysts. Anecdotally we recall the water-like viscosity of the straw-colored fluid contained in simple cysts, and how it obviously differs in viscosity from the thick brown chocolate-like material contained in endometriomas. Viscosity is a measurement of the flow properties of a substance. It is the ratio of the shear force applied, to the amount of resulting deformation or shear. Fluids are classified as Newtonian or non-newtonian depending on whether the rate of shear is directly proportional to the shearing force. Once we have a clearer understanding of how the three variables of pulse intensity, fluid attenuation and fluid viscosity affect the sonographic finding of acoustic streaming, we envisage that acoustic streaming could become an important adjunct to the routine sonographic examination of an adnexal cyst. In this study we used B-mode scanning to detect the presence of acoustic streaming. Our study s primary aim was to establish whether a difference exists in acoustic streaming characteristics in different adnexal cysts. We chose the objective outcome, the presence or absence of acoustic streaming, in the first instance. Previously, authors investigating acoustic streaming in other settings have used Doppler methods, not only to detect the presence of streaming, but also to measure its velocity 6 8,17. We have now begun work testing the ability of Doppler ultrasound to measure the velocity of streaming in adnexal cysts. We believe that this may further add to the value of acoustic streaming as an imaging tool.

78 Edwards et al. REFERENCES 1. Jermy K, Luise C, Bourne T. The characterization of common ovarian cysts in premenopausal women. Ultrasound Obstet Gynecol 2001; 17: 140 144. 2. Patel MD, Feldstein VA, Chen DC, Lipson SD, Filly RA. Endometriomas: Diagnostic performance of ultrasound. Radiology 1999; 210: 739 745. 3. Torr GR. The acoustic radiation force. Am J Phys 1984; 52: 402 408. 4. Wu J, Du G. Acoustic streaming generated by a focussed Gaussian beam and finite amplitude tone bursts. Ultrasound Med Biol 1993; 19: 167 176. 5. Starritt HC, Hoad CL, Duck FA, Nassiri DK, Summers IR, Vennart W. Measurement of acoustic streaming using magnetic resonance. Ultrasound Med Biol 2000; 26: 321 333. 6. Nightingale KR, Kornguth PJ, Walker WF, McDermott BA, Trahey GE. A novel ultrasonic technique for differentiating cysts from solid lesions: preliminary results in the breast. Ultrasound Med Biol 1995; 21: 745 751. 7. Nightingale KR, Kornguth PJ, Trahey GE. The use of acoustic streaming in breast lesion diagnosis: a clinical study. Ultrasound Med Biol 1999; 25: 75 87. 8. Shi X, Martin RW, Vaezy S, Kaczkowski P, Crum LA. Colour Doppler detection of acoustic streaming in a hematoma model. Ultrasound Med Biol 2001; 27: 1255 1264. 9. Fried AM, Rhodes RA, Morehouse IR. Endometrioma: analysis and sonographic classification of 51 documented cases. South Med J 1993; 86: 297 301. 10. Guerriero S, Mais V, Ajossa S, Paoletti AM, Angiolucci M, Melis GB. Tumor markers and transvaginal ultrasonography in the diagnosis of endometrioma. Obstet Gynecol 1996; 88: 403 407. 11. Jain KA. Prospective evaluation of adnexal masses with endovaginal gray-scale and duplex and color Doppler US: correlation with pathologic findings. Radiology 1994; 191: 63 67. 12. Volpi E, De Grandis T, Zuccaro G, La Vista A, Sismondi P. The role of transvaginal sonography in the detection of endometriomata. J Clin Ultrasound 1995; 23: 163 167. 13. Alcazar JL, Laparte C, Jurado M, Lopez-Garcia G. The role of transvaginal ultrasonography combined with color velocity imaging and pulsed Doppler in the diagnosis of endometrioma. Fertil Steril 1997; 67: 487 491. 14. Nyborg WLM. Acoustic streaming. In: Physical Acoustics, IIB, Mason WP (ed). Academic Press: New York, 1965; 265 331. 15. Frenkel V, Gurka R, Liberzon A, Shavit U, Kimmel. Preliminary investigations of ultrasound induced acoustic streaming using particle image velocimetry. Ultrasonics 2001; 39: 153 156. 16. HDI 5000 Ultrasound System Acoustic Output Tables. ATL Ultrasound 2000; Bothell, WA, USA. 17. Zauhar G, Starritt HC, Duck FA. Studies of acoustic streaming in biological fluids with an ultrasound Doppler technique. Br J Radiol 1998; 71: 297 302. SUPPLEMENTARY MATERIAL ON THE INTERNET The following material is available from the Journal homepage: http://www.interscience.wiley.com/jpages/0960-7692/suppmat/index.html (restricted access) Videoclip S1 Ultrasound video of the endometrioma shown in Figure 2 demonstrating the lack of acoustic streaming of the cyst particles when the color box was applied. Videoclip S2 Ultrasound video of the serous cystadenoma shown in Figure 4, demonstrating acoustic streaming. The particles within the cyst could be seen to move steadily away from the transducer when the color box was applied but their movement quickly ceased as soon as the color box was moved away.