Detection of tissue coagulation by decorrelation of ultrasonic echo signals in cavitation-enhanced high-intensity focused ultrasound treatment
© Yoshizawa et al. 2016
Received: 2 October 2015
Accepted: 4 April 2016
Published: 14 April 2016
A noninvasive technique to monitor thermal lesion formation is necessary to ensure the accuracy and safety of high-intensity focused ultrasound (HIFU) treatment. The purpose of this study is to ultrasonically detect the tissue change due to thermal coagulation in the HIFU treatment enhanced by cavitation microbubbles.
An ultrasound imaging probe transmitted plane waves at a center frequency of 4.5 MHz. Ultrasonic radio-frequency (RF) echo signals during HIFU exposure at a frequency of 1.2 MHz were acquired. Cross-correlation coefficients were calculated between in-phase and quadrature (IQ) data of two B-mode images with an interval time of 50 and 500 ms for the estimation of the region of cavitation and coagulation, respectively. Pathological examination of the coagulated tissue was also performed to compare with the corresponding ultrasonically detected coagulation region.
The distribution of minimum hold cross-correlation coefficient between two sets of IQ data with 50-ms intervals was compared with a pulse inversion (PI) image. The regions with low cross-correlation coefficients approximately corresponded to those with high brightness in the PI image. The regions with low cross-correlation coefficients in 500-ms intervals showed a good agreement with those with significant change in histology.
The results show that the regions of coagulation and cavitation could be ultrasonically detected as those with low cross-correlation coefficients between RF frames with certain intervals. This method will contribute to improve the safety and accuracy of the HIFU treatment enhanced by cavitation microbubbles.
KeywordsHIFU Cavitation Ultrasound imaging Coagulation detection Decorrelation
High-intensity focused ultrasound (HIFU) is a noninvasive technique for thermal ablation of solid tumors and has already been used to treat fibroids and prostatic tumors. Ultrasound can be focused to a target tumor such as cancer to be thermally coagulated selectively . However, unlike in an open surgery, the operator or surgeon cannot look at the tissue being or to be treated by a naked eye in real time. Therefore, a noninvasive technique to monitor thermal lesion formation is necessary to ensure the accuracy and safety of HIFU treatment.
Magnetic resonance imaging (MRI) and ultrasonic imaging are currently used for monitoring HIFU treatment and its therapeutic effects. MRI has the advantage in tissue temperature monitoring , but it lacks real-time monitoring capability especially for cavitation. Ultrasonic imaging is chosen in this study because of its higher spatial and temporal resolution at a much lower cost. A hyper-echoic change in a normal B-mode image at the focal point of HIFU has been used for estimating the coagulation area, but the change is faint unless bubbles are generated due to cavitation or boiling . Slight changes in ultrasonic backscatter were observed corresponding to thermal coagulation as well as changes in the speed of sound and attenuation coefficient in a liver tissue . In hepatic tissue, the relation between the change in ultrasonic backscatter and the slight increase in the hepatic cell concentration was shown, which was observed in the histological structure . Taking this slight change into account, the method to estimate the coagulation by comparing two radio-frequency (RF) frames which correspond to the two B-mode images during HIFU exposure using the distribution of cross-correlation coefficient has been studied, and the decorrelation corresponding to coagulation was observed in the focal area of HIFU exposure even in cases that bubbles due to boiling or cavitation were not observed [6, 7]. The echo decorrelation imaging has been investigated also for the ultrasound monitoring of RF ablation .
With a typical equipment for current HIFU treatment, an ultrasound exposure at 4 MHz for 3 s can treat a volume of 2 × 2 × 10 to 3 × 3 × 12 mm3 , resulting in that the median operating time of 142 min (ranging 35–390 min), including cooling time between consecutive HIFU exposures, is needed to treat the median prostate volume of 21.9 cm3 (ranging 4.6–68.8 cm3) . This long treatment time is thought to be a significant drawback of HIFU treatment. The enhancement of ultrasonic heating by cavitation bubbles is receiving a lot of attention to solve this problem [11–16]. Acoustic cavitation is the phenomenon in which microbubbles in the order of 0.1–10 μm in diameter are generated by acoustic irradiation. In HIFU treatment, cavitation bubbles can be generated and vibrated by ultrasound, can enhance the heating effect, and make the treatment time much shorter [17–21].
The objective of this study is to develop a noninvasive technique to monitor the thermal lesion formation, even in the existence of such cavitation bubbles. In this study, high-speed imaging by parallel beamforming is performed using ultrasound RF signals acquired during HIFU exposure, and the cross-correlation coefficient between the obtained RF frames is calculated according to the previous study . Then, both coagulation and cavitation regions are detected as the regions of low correlation. They should be distinguished by changing the interval between the RF frames for cross-correlation because the order of the time constant is quite different for cavitation bubbles and coagulation.
HIFU exposure and RF data acquisition sequence
In this study, RF signals were obtained by combining two imaging ultrasound irradiation methods. First, plane wave transmission followed by parallel beamforming for high-speed ultrasonic imaging, at a frame rate higher than 5000 fps , was applied. To improve the contrast ratio of the B-mode image, multiple steered plane waves were transmitted, and the obtained ultrasonic images were coherently compounded [23, 24]. To visualize cavitation bubbles, “pulse inversion (PI)”  sequence was employed. Plane wave transmission was performed six times in total, with a separation angle of 6°, at −6°, 0°, and +6°. At each angle, twice transmission of one-cycle sine wave with opposite phases was performed. The phases of the first and second sine waves were 0° and 180°, respectively. The pulse repetition period of the six plane waves was 0.2 ms. The transmission center frequency and the sampling frequency were set at 4.5 and 36 MHz, respectively. These parameters may not be thoroughly optimized, so further optimization will be possible if necessary. The ultrasonic imaging plane was set so that it contained the axis of the HIFU beam as seen in Fig. 1. The HIFU focal point was 50.0 mm from the probe surface in the depth direction and 12.5 mm from the left edge of the B-mode images in the lateral direction. The peak-to-peak pressure of the transmitted plane waves was 2.0 MPa with a steering angle of 0° at a depth of 50.0 mm.
A programmable ultrasound imaging system (V-1 System, Verasonics, WA, USA) with a linear array probe (UST-5412, Hitachi Aloka Medical, Tokyo, Japan) was used for ultrasonic monitoring. The data of 200 RF frames were acquired at a frame rate of 20 fps, resulting in a total duration of 10 s. To avoid the echoes of HIFU, the first plane wave transmission was performed 0.3 ms after the heating waves stopped as seen in Fig. 2.
Cross-correlation coefficient distribution
The maximum value of the correlation coefficient described by Eq. (2) was plotted as a function of the reference window position. Then, cross-correlation coefficients were calculated for the next reference window, which 50 % overlapped with the previous one for a sufficient spatial resolution. A motion-compensated distribution of correlation coefficient between the two entire images can thus be depicted. Both coagulation and cavitation regions were estimated by changing the interval between the RF frames for cross-correlation because the orders of magnitude of time constants of cavitation bubbles and tissue coagulation are quite different. In this study, the intervals to detect cavitation and coagulation were set to 50 and 500 ms, respectively.
After a series of HIFU exposure and RF signal acquisition, samples were observed pathologically. They were frozen by use of liquid nitrogen within 5 h and cut in 10-μm-thick slices in the direction parallel to the imaging plane. A slice was obtained every 100 μm perpendicular to the plane, and the slice with the largest coagulation region was chosen and stained with hematoxylin and eosin (H&E). The muscle fiber atrophy due to thermal coagulation was observed , so the coagulation region was determined as a coarse region of atrophic muscle fibers in a low-power microscopic field.
Estimation of the cross-correlation coefficient threshold
Distribution of cross-correlation coefficient with 50-ms intervals
Distribution of cross-correlation coefficient with 500-ms intervals
Cross-correlation coefficient threshold
Size of decorrelated area
Comparison between distribution of cross-correlation coefficient with 50-ms intervals and pulse inversion image
Comparison between distribution of cross-correlation coefficient with 500-ms intervals and histology
The comparison between the observed and estimated coagulation area
Observed coagulation area in histology, A HE [mm2]
Estimated coagulation area, A US [mm2]
Cross-correlation coefficient threshold for 500-ms intervals
Ratio of A US to A HE (%)
Comparison between distributions of cross-correlation coefficient with 50- and 500-ms intervals
Figure 8b demonstrates that the correlation coefficient particularly decreased with 50-ms intervals at the focal spot of HIFU with a sequence inducing cavitation while it hardly decreased with a sequence inducing no cavitation. Furthermore, Fig. 12 shows that the distribution of minimum hold cross-correlation coefficients agrees with the corresponding PI image. As the trigger pulses were exposed every 50 ms, RF echo from cavitation bubbles could change significantly in 50 ms, which should cause the decrease of the cross-correlation coefficients. Therefore, it is necessary to detect the decorrelation area with 500-ms intervals outside the area with 50-ms intervals as shown in Fig. 12a for the coagulation detection when RF echo from cavitation bubbles cannot be neglected.
The decrease in correlation coefficient with 500-ms intervals has a potential to be used for estimating the region of tissue where coagulation occurs even when cavitation is employed, because of the following reasons. First, the size of the estimated areas agreed well with the coagulation areas observed in histology as shown in Table 1. Second, all the areas regarded as those of cavitation are contained in the estimated coagulation areas as shown in Fig. 13. Therefore, the proposed way of estimating a coagulation area does not seem to significantly suffer from the decrease in correlation due to cavitation, at least, in the exposure sequences as tested.
The coagulation area from histology of sample (1) in Table 1 contains an area where the muscle fibers are a little denser than other coagulated areas, which can be seen in lower right area in Fig. 6a. The temperature rise in the denser area might be lower in other coagulated areas. The denser area approximately corresponds to that around a depth and width of 50 and 6 mm, respectively, in Fig. 13a, where the correlation coefficients are larger than the thresholds for 50- and 500-ms interval data. As a result, the denser area in Fig. 6a was included in A HE but not in A US as shown in Fig. 13a. If such an area in A HE is eliminated from the calculation, the ratio, A US/A HE, will become to 98 %, much closer to 100 %. The decorrelation detection using histograms in this study is based on the detection of the significant change in tissue structure due to coagulation. If there are no significant changes in tissue structure, it should be difficult to detect. Therefore, the method may underestimate the treated region. It may be suitable for monitoring to ensure the therapeutic effects of HIFU thermal treatments but may not be sufficient for monitoring the side effects. Also, the method may be applicable to the monitoring of lesion formation in histotripsy, where tissues are mechanically fractionated and the tissue changes probably cause the decorrelation of the ultrasound echo signals.
Figure 13b shows a decorrelated area ranging in a depth of 54–56 mm, and a width of 3.5–10-mm point is not in the HIFU focal zone. This decorrelated region is thought to have been due to the motion of preexisted bubbles. Figure 13c, d also shows decorrelated areas below the HIFU focal zone. These decorrelated regions are thought to have been caused by a kind of acoustic shadows of the cavitation bubbles in the HIFU focal zone.
Here, the shape of the coagulation volume is assumed to be a spheroid and the axis of symmetry is parallel to the propagation of HIFU. The coagulation volume estimated from the H&E-stained tissue sample was 92 mm3. Although the estimated coagulation volume from the decorrelation shows a good agreement with that estimated from H&E-stained tissue sample in the case of Fig. 13b, the reasonable volume estimation seems to be difficult especially in the case of Fig. 13c, d because of the complicated shape of the decorrelated areas.
Temporal change in estimated areas
I. The estimated area with 50-ms intervals increased quickly for the first 3 s of the exposure and gradually after that. This is interpreted that the cavitation bubbles were generated only near the HIFU focal point for the first 3 s, and then they started being generated in the area migrating toward the HIFU transducer .
II. The estimated area with 500-ms intervals increased even more quickly than that with 50-ms intervals from 1 to 4 s after the start of the exposure. This is interpreted that tissue coagulation started and expanded near the focal point in this time zone.
III. The estimated area with 500-ms intervals increased gradually at a similar slope as with 50-ms intervals from 4 to 8 s after the start of the exposure. This is interpreted that they increased similarly only due to cavitation for a while after the rapid coagulation near the focal point has stopped.
IV. The estimated area with 500-ms intervals increased quickly from 8 s after the start of the exposure. This is interpreted that coagulation started a little distance away from the focal point. For example, the decorrelation area around a depth and width of 50 and 5 mm in Fig. 13a was observed in this time zone, as shown in Fig. 9c, while the decrease in correlation coefficients to below the threshold with 50-ms intervals are not particularly observed in the area as shown in Fig. 7c. This result suggests that a part of the decorrelation areas with 500-ms intervals outside those with 50-ms intervals could be detected without the influence of the decorrelation due to cavitation bubbles, as seen in this time zone, resulting in a good agreement between the observed and estimated coagulation area.
In this study, ultrasonic RF signals during HIFU exposure were used to estimate the cavitation and coagulation regions by calculating cross-correlation coefficients between IQ data of RF frames with certain intervals. The results show that the coagulation region was approximately estimated from the decorrelation map with 500-ms intervals and the size of decorrelated area in the map matched that of the actual coagulation area observed in histology with a cross-correlation coefficient threshold set properly. The estimation of coagulation regions with 500-ms intervals outside decorrelated regions with 50-ms intervals was not significantly affected by cavitation preceding the coagulation, at least in a certain HIFU sequence. This method for the coagulation detection will contribute to improve the safety and accuracy of the HIFU treatment enhanced by cavitation microbubbles.
This study was partly supported by grant-in-aid from the Ministry of Health, Labour and Welfare. The authors would like to express their sincere thanks to Tetsuya Kodama and Yukari Yanagisawa (Tohoku University, Japan) for their excellent support with the histology study and Satoshi Tamano (Hitachi Aloka Medical, Japan) for his cooperation in constructing the ultrasound imaging setup.
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. The Creative Commons Public Domain Dedication waiver (http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
- Ter Haar GR. Ultrasound focal beam surgery. Ultrasound Med Biol. 1995;21:1089.View ArticlePubMedGoogle Scholar
- Tempany CM, Stewart EA, McDannold N, Quade BJ, Jolesz FA, Hynynen K. MR Imaging–guided focused ultrasound surgery of uterine leiomyomas: a feasibility study. Radiology. 2003;226:897.View ArticlePubMedGoogle Scholar
- Rabkin BA, Zderic V, Vaezy S. Hyperecho in ultrasound images of HIFU therapy: involvement of cavitation. Ultrasound Med Biol. 2005;31:947.View ArticlePubMedGoogle Scholar
- Bush NL, Rivens I, Ter Haar GR, Bamber JC. Acoustic properties of lesions generated with an ultrasound therapy system. Ultrasound Med Biol. 1993;19:789.View ArticlePubMedGoogle Scholar
- Shishitani T, Matsuzawa R, Yoshizawa S, Umemura S. Changes in backscatter of liver tissue due to thermal coagulation induced by focused ultrasound. J Acoust Soc Am. 2013;134:1724.View ArticlePubMedGoogle Scholar
- Matsuzawa R, Shishitani T, Yoshizawa S, Umemura S. Monitoring of lesion induced by high-intensity focused ultrasound using correlation method based on block matching. Jpn J Appl Phys. 2012;51:07GF26.View ArticleGoogle Scholar
- Sasaki S, Takagi R, Matsuura K, Yoshizawa S, Umemura S. Monitoring of high-intensity focused ultrasound lesion formation using decorrelation between high-speed ultrasonic images by parallel beamforming. Jpn J Appl Phys. 2014;53:07KF10.View ArticleGoogle Scholar
- Mast TD, Pucke DP, Subramanian SE, Bowlus WJ, Rudich SM, Buell JF. Ultrasound monitoring of in vitro radio frequency ablation by echo decorrelation imaging. J Ultrasound Med. 2008;27:1685.PubMedGoogle Scholar
- Uchida T, Sanghvi NT, Gardner TA, Koch MO, Ishii D, Minei S, Satoh T, Hyodo T, Irie A, Baba S. Transrectal high-intensity focused ultrasound for treatment of patients with stage T1b-2n0m0 localized prostate cancer: a preliminary report. Urology. 2002;59:394.View ArticlePubMedGoogle Scholar
- Uchida T, Shoji S, Nakano M, Hongo S, Nitta M, Murota A, Nagata Y. Transrectal high-intensity focused ultrasound for the treatment of localized prostate cancer: eight-year experience. Int J Urol. 2009;16:881.View ArticlePubMedGoogle Scholar
- Holt RG, Roy RA. Measurements of bubble-enhanced heating from focused. MHz-frequency ultrasound in a tissue-mimicking material. Ultrasound Med Biol. 2001;27:1399.View ArticlePubMedGoogle Scholar
- Bailey MR, Couret LN, Sapozhnikov OA, Khokhlova VA, Gter H, Vaezy S, Shi X, Martin R, Crum LA. Use of overpressure to assess the role of bubbles in focused ultrasound lesion shape in vitro. Ultrasound MedBiol. 2001;27:695.View ArticleGoogle Scholar
- Sokka SD, King R, Hynynen K. MRI-guided gas bubble enhanced ultrasound heating in in vivo rabbit thigh. Phys Med Biol. 2003;48:223.View ArticlePubMedGoogle Scholar
- Kaneko Y, Maruyama T, Takegami K, Watanabe T, Mitsui H, Hanajiri K, Nagawa H, Matsumoto Y. Use of a microbubble agent to increase the effects of high intensity focused ultrasound on liver tissue. Eur Radiol. 2005;15:1415.View ArticlePubMedGoogle Scholar
- Umemura S, Kawabata K, Sasaki K. In vivo acceleration of ultrasonic tissue heating by microbubble agent. IEEE Trans Ultrason Ferroelectr Freq Control. 2005;52:1690.View ArticlePubMedGoogle Scholar
- Kawabata K, Sugita N, Yoshikawa H, Azuma T, Umemura S. Nanoparticles with multiple perfluorocarbons for controllable ultrasonically induced phase shifting. Jpn J Appl Phys. 2005;44:4548.View ArticleGoogle Scholar
- Takagi R, Yoshizawa S, Umemura S. Enhancement of localized heating by ultrasonically induced cavitation in high intensity focused ultrasound treatment. Jpn J Appl Phys. 2010;49:07HF21.Google Scholar
- Inaba Y, Yoshizawa S, Umemura S. Coagulation of large regions by creating multiple cavitation clouds for high intensity focused ultrasound treatment. Jpn J Appl Phys. 2010;49:07HF22.Google Scholar
- Inaba Y, Moriyama T, Yoshizawa S, Umemura S. Ultrasonic coagulation of large tissue region by generating multiple cavitation clouds in direction perpendicular to ultrasound propagation. Jpn J Appl Phys. 2011;50:07HF13.View ArticleGoogle Scholar
- Sasaki H, Yasuda J, Takagi R, Miyahsita T, Goto K, Yoshizawa S, Umemura S. Highly efficient cavitation-enhanced heating with dual-frequency ultrasound exposure in high-intensity focused ultrasound treatment. Jpn J Appl Phys. 2014;53:07KF11.View ArticleGoogle Scholar
- Goto K, Takagi R, Miyashita T, Jimbo H, Yoshizawa S, Umemura S. Effect of controlled offset of focal position in cavitation-enhanced high-intensity focused ultrasound treatment. Jpn J Appl Phys. 2015;54:07HF12.View ArticleGoogle Scholar
- Montaldo G, Tanter M, Bercoff J, Benech N, Fink M. Coherent plane-wave compounding for very high frame rate ultrasonography and transient elastography. IEEE Trans Ultrason Ferroelectr Freq Control. 2009;56:489.View ArticlePubMedGoogle Scholar
- Berson M, Roncin A, Pourcelot L. Compound scanning with an electrically steered beam. Ultrason Imaging. 1981;3:303.View ArticleGoogle Scholar
- Entrekin RR, Porter BA, Sillesen HH, Wong AD, Cooperberg PL, Fix CH. Real-time spatial compound imaging: application to breast, vascular, and musculoskeletal ultrasound. Semin Ultrasound CT MR. 2001;22:50.View ArticlePubMedGoogle Scholar
- Simpson DH, Chin CT, Burns PN. Pulse inversion Doppler: a new method for detecting nonlinear echoes from microbubble contrast agents. IEEE Trans Ultrason Ferroelect Freq Control. 1999;46:372.View ArticleGoogle Scholar
- Tornberg E. Effects of heat on meat proteins—implications on structure and quality of meat products. Meat Sci. 2005;70:493.View ArticlePubMedGoogle Scholar
- Maxwell AD, Wang T-Y, Cain CA, Fowlkes JB, Sapozhnikov OA, Bailey MR, Xu Z. Cavitation clouds created by shock scattering from bubbles during histotripsy. J Acoust Soc Am. 2011;130:1888.View ArticlePubMedPubMed CentralGoogle Scholar