Photoacoustic Detection and Optical Spectroscopy of High-Intensity Focused Ultrasound-Induced Thermal Lesions in Biologic Tissue

.................................................................................................iii ACKNOWLEDGMENTS..................................................................................iv TABLE OF CONTENTS..................................................................................vi LIST OF SYMBOLS AND ABBREVIATIONS.....................................................viii LIST OF TABLES...........................................................................................ix LIST OF FIGURES............................................................................................x LIST OF APPENDICES..................................................................................xii Chapter 1: Introduction......................................................................................1 1.1 Therapeutic ultrasound...................................................................................1 1.2 High-intensity focused ultrasound (HIFU)............................................................3 1.2.1 HIFU therapeutic mechanisms....................................................................3 1.2.1.1 Thermal mechanism of HIFU.............................................................3 1.2.1.2 Mechanical mechanism of HIFU..........................................................4 1.2.2 Image-guided HIFU................................................................................5 1.2.3 Limitation of current HIFU imaging modalities................................................6 1.3 Photoacoustics and its potential in thermal therapy...................................................6 1.3.1 Principles of photoacoustics.........................................................................6 1.3.2 The potential of photoacoustic detection of HIFU treatments...............................8 1.4 Optical spectroscopy of biologic tissues...............................................................9


INTRODUCTION
High-intensity focused ultrasound (HIFU) is a bloodless surgical modality that is capable of inducing thermal and mechanical effects in deep-seated regions of interest (ROI) selectively and noninvasively. 1 Using a high-power focused transducer, HIFU beams can be achieved and focalized within the transducer's focal zone in order to deposit acoustic energy in the ROI with minimal or no harm to intervening tissue layers.Within the ROI, the selectively deposited acous-tic energy is converted into heat, raising the temperature of the targeted tissue volume, while generally keeping the temperature of the surrounding tissue volumes at physiologically safe levels.The increase in the targeted tissue temperature above 56 • C for at least 1 s can lead to immediate cell death due to protein denaturation in a process known as coagulation necrosis. 2The spatially confined, coagulated tissue volumes are termed thermal lesions, 3 since the predominant mechanism of HIFU is thermal, although cavitation and boiling bubbles are also involved in the formation and shaping of the induced thermal lesions, 2 particularly at higher temperatures.
Currently, most HIFU treatments are monitored and guided with magnetic resonance imaging (MRI) (Refs.11, 13, 14, 17,  and 19-23) and ultrasound imaging. 24,25 owever, the challenges associated with the aforementioned imaging modalities, including cost and equipment compatibility with MRI as well as low contrast-resolution between treated and normal tissues with ultrasound imaging, have stimulated the interest of many research groups to investigate alternative HIFU monitoring tools.Among the proposed techniques is photoacoustic (PA) imaging.It is based on the optical absorption of a short, pulsed laser irradiation and the subsequent generation of acoustic waves.These laser-induced acoustic waves, generated as a result of thermoelastic expansion in the absorbing medium, are detectable by a diagnostic ultrasound transducer. 26Therefore, PA imaging combines the advantages of both ultrasound and optical imaging modalities, i.e., it provides higher resolution than optical imaging and higher tissue contrast than ultrasound imaging. 26The higher resolution of PA imaging is due to the weaker ultrasound scattering in tissues than optical scattering. 26On the other hand, the higher tissue contrast of PA imaging is due to its capability in visualizing the tissue optical absorption properties, providing the advantage of functional imaging of important physiological parameters such as hemoglobin concentration. 26Moreover, PA imaging is a noninvasive modality that utilizes a nonionizing electromagnetic radiation.
The possibility of PA detection of HIFU-induced thermal lesions has been demonstrated in vitro, 27,28 ex vivo, 29 and in vivo. 30Moreover, previous studies have also been successful in detecting laser-induced thermal lesions photoacoustically. 31,32 ecause the generated PA pressure depends on the absorbed optical energy (i.e., the product of laser energy fluence and optical absorption coefficient) and the thermoacoustic efficiency of the absorbing structure, a change in at least one of these factors can result in a change in the detected PA signal.Therefore, thermally induced changes in the optical [33][34][35][36] and thermomechanical 37,38 properties of tissues have been investigated following laser-induced thermotherapy or water/saline-bath heating.However, to our knowledge, the optical and thermomechanical properties of HIFU-treated tissues have not yet been examined in conjunction with PA detection of HIFU treatments.In this work, we demonstrate the possibility of PA detection of HIFU-induced thermal lesions in chicken breast tissues in vitro at 720 and 845 nm laser illuminations.Moreover, for the first time, we measure the optical properties (total transmittance, diffuse reflectance, absorption coefficient, reduced scattering coefficient, effective attenuation coefficient, and light penetration depth) of HIFU-induced thermal lesions and compare them with native (untreated) tissues in the wavelength range of 500-900 nm.The aforementioned objectives are intended to test our hypothesis that PA detection of HIFU-induced thermal lesions is feasible due, in part, to changes in the optical properties of HIFU-treated tissues.

2.A. The HIFU transducer and its driving electronics
A single-element, spherically concaved HIFU transducer (6699A101, Imasonic S. A.,Voray-sur-l'Ognon, France) made of a high-power piezocomposite material was used in this study.The transducer had a resonance frequency of 1 MHz, aperture diameter of 125 mm, and radius of curvature of 100 mm.It was calibrated using a needle hydrophone (HNA-0400, ONDA Corporation, Sunnyvale, CA) as well as a radiation force balance unit (RFB-2000, ONDA Corporation, Sunnyvale, CA). 39The transducer's efficiency (ratio of measured output acoustic power to the net input electric power) was determined to be ∼0.64 (Ref.39) for the net input electric power used in this study, which was ∼110 W. Based on a numerical simulation of the acoustic field, the HIFU transducer had a depth of field of ∼7.55 mm and a −6 dB focal beam width of 1.24 mm.
The HIFU transducer was driven by an arbitrary function generator (AFG3022B, Tektronix, Beaverton, OR), which generated electrical radiofrequency (RF) signals.An RF power amplifier (A150, Electronics and Innovation, Rochester, NY) was utilized to amplify the generated RF signals.To maximize the transmitted power to the HIFU transducer, a matching network connected the RF power amplifier with the transducer.Moreover, a multilogger thermometer (HH506RA, OMEGA Engineering Inc., Stamford, CT) was used to monitor the change in the HIFU transducer's surface temperature during treatments, thereby ensuring safe operation of the transducer.A schematic diagram of the experimental setup is shown in Fig. 1.

2.B. HIFU treatments
As a biologic tissue model for HIFU treatments, in vitro chicken breast tissues, purchased from a local market, were used in this study.For both the PA detection and optical spectroscopy experiments, the tissues were cut into approximately 7.5 cm × 6.5 cm × 1.5 cm (length × width × thickness) samples, and five HIFU thermal lesions were created in each sample by manually moving the tissue holder using a micropositioning controller (Fig. 1) to five different locations separated by 5 mm.The sequence of generating the five HIFU thermal lesions is illustrated in Fig. 2(a).The treated region in each sample, which consisted of five HIFU thermal lesions, is shown in Fig. 2(b).
Operated with 60% duty cycle, the HIFU transducer generated focused waves that propagated through ∼9.0 cm in 053502-3 FIG. 1. Schematic diagram of the experimental setup of HIFU treatments.The focus of the HIFU transducer was at 10.0 cm.The distance between the centre of the transducer and the transparent plastic sheet covering the opening of the tissue holder was ∼9.0 cm, resulting in a tissue focal depth of ∼1.0 cm.degassed water at room temperature (∼20 • C) and 1.0 cm in tissue, creating thermal lesions with acoustic intensity, I SPTA , of ∼1690 W/cm 2 at the transducer's focus inside tissue.The free-field I SPTA was ∼1910 W/cm 2 .The transmitted acoustic power in each HIFU treatment was ∼42 W, and the total sonication time was 30 s.

2.C.1. The PA system
PA detection of HIFU treatments was achieved using the Imagio Small Animal PA imaging system (Seno Medical Instruments Inc., San Antonio, TX).A schematic diagram of the experimental setup is shown in Fig. 3.
The utilized PA imaging system consisted of a pulsed titanium:sapphire (Ti:Al 2 O 3 ) laser with wavelengths tunable in the range between 720 and 850 nm.Delivered through an articulated arm, the pulsed laser had a 9-mm beam diameter,  6-ns pulse width, and 10-Hz pulse repetition rate.The maximum output energy of the laser at the wavelengths used in this study was ∼128 mJ, which corresponds to a maximum laser energy fluence of 201 mJ/cm 2 at the irradiated tissue surface.This energy level causes a thermoelastic expansion and results in strong PA signals from the native and treated chicken breast tissues, although they do not inherently contain very strong light chromophores.
The photoacoustic receiver of the PA imaging system was a 4-element annular array ultrasound transducer with a centre frequency of 5 MHz.The −6 dB bandwidth of the transducer was 60% and its focal length was 29.5 mm, as determined previously. 40The transducer, which had a depth of field of ∼10 mm, was coaxially mounted at a distance of 105 mm from the laser source in order to allow raster scanning of the chicken breast sample under investigation, which was placed within the transducer's depth of field (Fig. 3).A 52-μm-thick, PVDF-based gold foil (210037022, Measurement Specialities Inc., Hampton, VA) was used to protect the ultrasound transducer from direct laser exposure.Moreover, a chargecoupled device (CCD) camera, placed underneath the ultrasound transducer, was utilized to facilitate targeting of the ROI by the laser beam by providing a real-time view of the target region that is at the cross hair of the laser.The CCD camera is part of the commercial Imagio PA system.

2.C.2. Tissue sample preparation for PA detection
Prior to performing the PA detection experiments, the HIFU-treated chicken breast samples were cut into approximately 4.0 cm × 3.0 cm (width × length) and were sliced to ∼0.5 cm thickness using a professional food slicer (1060-C, The Rival Company, El Paso, TX).Slicing the chicken breast samples eliminated the undesired skin part of the tissue, and reduced the extension of the superficial thermal lesion in the sliced tissue volume to ∼0.35 cm.Moreover, in order to prevent contamination of the water tank, the samples were plastic wrapped and vacuum sealed with a vacuum sealer (V2060 FoodSaver, Jarden Corp., Rye, NY).

2.C.3. PA measurements
The plastic-wrapped, vacuum-sealed samples, shown in Fig. 2 within the focal zone of the ultrasound transducer inside the PA imaging system's water reservoir (Fig. 3).To compare the PA response of HIFU-treated and native tissues, two ROIs were identified in each one of the seven samples examined in this study.One of the ROIs was within a HIFU-treated tissue volume, while the other ROI was within a native tissue volume in the sample under investigation.Horizontal raster scanning was performed with 0.1 mm scan resolution covering 11 locations within each ROI.The laser fired four pulses at each location in the ROI and the detected PA RF signals, which were collected within 0.4 s, were averaged to generate a single PA RF line.Therefore, a total of 11 PA RF lines, obtained from 11 locations, were recorded for each ROI within each sample.The acquired PA data were imported to MAT- LAB (R2012b, The MathWorks Inc., Natick, MA) for processing.For each one of the seven samples, the 11 PA RF lines were averaged, producing a single PA RF line that corresponds to a particular ROI (treated or native) within that sample.This averaging method was performed since the samples were well-controlled and stable, i.e., they did not produce large variations in space or time within the examined regions.The PA detection experiments were performed with two laser wavelengths: 720 and 845 nm, thereby covering the maximum range of wavelengths generated by our PA imaging system.
It is noteworthy to mention that the PA detection experiments were conducted after assessing the accuracy of the CCD camera in targeting the position of an ROI (Fig. 3) using carbon rods, which are known to produce very strong PA signals.Based on the strength of the acquired PA signals, this testing allowed us to make subtle adjustments in the position of the CCD camera so that the ROI, shown in the real-time CCD camera's image, was coincident with the area that was irradiated by the laser beam.

2.D. Optical spectroscopy
Optical spectroscopy experiments were conducted in this study in order to understand the influence of HIFU-induced changes in the optical properties on the detected PA pressure rise, P(z).The PA pressure rise is directly proportional to the absorbed optical energy and the thermoacoustic efficiency, and is given in the equation below: 41,42 where is the dimensionless Grüneisen parameter, which represents the fraction of thermal energy that is converted into acoustic waves (also known as thermoacoustic efficiency).In Eq. ( 1), the absorbed optical energy from a laser source, E abs (z) [J/m 3 ], is the product of the optical absorption coefficient, μ a [m −1 ], and the laser energy fluence, (z) [J/m 2 ].
Our spectroscopic investigation included measurements of diffuse reflectance (DR) and total transmittance (TT) of chicken breast tissue slices, prepared as described in Sec.2.D.1, using a commercially available spectrophotometer with an integrating sphere attachment (Sec.2.D.2).The inverse adding-doubling method, described in Sec.2.D.3, was utilized to determine the optical absorption coefficient, μ a , and the reduced scattering coefficient, μ s , from the experi-mental data of the DR and TT.The effective attenuation coefficient, μ eff , and the effective light penetration depth, D eff , were calculated in the wavelength range of 500-900 nm using the following equations: 41 (3)

2.D.1. Tissue sample preparation for optical spectroscopy
A new set of chicken breast tissues treated with HIFU, as described in Sec.2.B, were stored at low temperatures (6 -8 • C) for 12-48 h prior to conducting optical spectroscopy experiments.Using the food slicer mentioned in Sec.2.C, the tissues were sliced to thin slices; each containing HIFUtreated and native tissue regions.The sliced tissue specimens were further trimmed with a knife to fit within a 1-mm spacing between two 1-mm-thick optical glass slides (BK7, Esco Optics Inc., Oak Ridge, NJ), which were spaced apart using stainless steel shims.Sandwiching sliced tissue specimens within the two optical glass slides reduced the tissue surface irregularities.The tissue specimen and the two optical glass slides covering it had a combined thickness of ∼3 mm, as measured using a micrometer.Figure 4 is a photograph of one of the 10 sliced tissue specimens that were prepared during our optical spectroscopy experiments.

2.D.2. Integrating-sphere measurements
To compare the optical properties of HIFU-treated and native tissues, the sliced tissue specimens, which were positioned between two optical glass slides, were used in optical spectroscopy experiments using the Shimadzu spectrophotometer (UV 3600 UV-VIS-NIR, Shimadzu Scientific Instruments, Columbia, MD) with the ISR-3100 integrating sphere attachment.The system's sphere diameter was 60.0 mm, entrance port diameter was 16.7 mm, and the exit port diameter was 18.8 mm.The integrating-sphere method allowed us carry out the DR and TT measurements in a dual-beam mode, with a single-integrating-sphere configuration.The utilized slit width, which was fixed during the experiments, was 8 nm, yielding a measured illumination beam of approximately 7.1 mm at the sample surface.The wavelength of grating and detector changes was set at 910 nm.The experimental setup of the DR and TT measurements is depicted in Fig. 5, which was reproduced, with permission, from Shimadzu Corporation. 43,44 total of 10 sliced tissue specimens were prepared as described in Sec.2.D.1.For each one of the 10 sliced tissue specimens, the DR and TT of native and HIFU-treated portions of the tissue were measured three times between 500 and 900 nm in 1-nm increments.Therefore, for a given sliced tissue specimen, four spectra were generated, each of which was an average of three trials.These spectra were: (a) TT and DR for the native portion of a specimen (two spectra), and (b) TT and DR for the HIFU-treated portion of a specimen (two spectra).

2.D.3. Inverse adding-doubling method
The inverse adding-doubling (IAD) program 45,46 was employed in this study to determine the optical absorption and the reduced scattering coefficients of HIFU-treated and native chicken breast tissues using the DR and TT data, collected as described in Sec.2.D.2.In the IAD program, optical parameters such as scattering, absorption, and scattering anisotropy can be determined by repeatedly solving the radiative transport equation (RTE) until a match with the measured reflection and transmission values is made. 45The "inverse" implies a reversal to the regular extraction of transmission and reflection values from the optical properties, whereas "adding-doubling" refers to the utilized methodology in solving the RTE. 45The IAD is a versatile method, and is wellsuited for measurements employing biologic tissue specimens positioned between optical glass slides. 45

2.E. Monte Carlo simulation of light distribution
The light transport in chicken breast tissue was simulated using a Monte Carlo (MC) method developed by Boas et al. 47 and modified by Saiko et al. 48.To precisely simulate the light distribution in the tissue, a total of 1 × 10 8 photons were launched at 720 nm.The tissue was modeled as a box with dimensions 40 mm × 30 mm × 5 mm, while the HIFU-induced thermal lesion was modeled as a cylinder with a diameter of 10 mm and a length of 3.5 mm.The MC simulation allowed us to estimate the light fluence in a native chicken breast tissue slab (i.e., a slab without a lesion) and in another chicken breast tissue slab containing a HIFU-induced thermal lesion, using the measured values of the reduced scattering and absorption coefficients (Fig. 10) at 720 nm.The anisotropy factor, g, was kept constant throughout the simulation (g = 0.9). 49he scattering coefficient, μ s = μ s /(1 − g), values used in the MC simulation for a native chicken breast tissue and a HIFU-induced thermal lesion were 2.95 and 39.06 mm −1 , respectively.The optical absorption coefficient values, which were extracted directly from Fig. 10(a), used in the MC simulation for a native chicken breast tissue and a HIFU-induced thermal lesion were 0.10 and 0.12 mm −1 , respectively.The laser beam diameter used in the simulation was 9 mm.

3.A. PA detection of HIFU-induced thermal lesions
Figures 6 and 7 show the PA RF signals of seven samples at 720 and 845 nm laser illumination, respectively, averaged from 11 locations within each ROI in the HIFU-treated and native tissue volumes of each sample.The numerical values that appear inside the rectangular boxes placed within each one of the 14 PA RF lines of Figs. 6 and 7 represent the ratio of the peak-to-peak PA signal amplitudes of HIFU-treated and native ROIs of their respective sample-wavelength combination.Combining the ratios of the seven samples investigated at 720 nm, the averaged ratio of the peak-to-peak PA signal amplitude of HIFU-treated tissue to that of native tissue is 3.68 ± 0.25 (mean ± SEM).At 845 nm, the averaged ratio of the peak-to-peak PA signal amplitude of HIFU-treated tissue to that of native tissue is 3.75 ± 0.26 (mean ± SEM). Figure 8 provides a comparison of the averaged ratios of the HIFU-treated and native tissue peak-to-peak PA signal amplitudes from the seven samples at 720 and 845 nm laser illumination.The increase in the PA signal amplitude by more than three times as a result of HIFU-induced coagulation has also been reported in a previous study. 28n Figs. 6 and 7, the slight shift and, in some cases, the differences in the width of the N-shapes of the PA signals of native and HIFU-treated tissues can be attributed to minor nonuniformity in the surface of the tissue samples used in the PA detection experiments.In few instances, the HIFUinduced thermal lesion appeared as a "bulging mass" that is thicker than the surrounding native tissue, making some of the N-shapes in Figs. 6 and 7 to appear of slightly different widths in native tissues compared to HIFU-induced thermal lesions.It is also important to note that HIFU-induced thermal lesions have a higher acoustic attenuation coefficient than native tissues. 50,51 herefore, laser-generated PA pressure waves should undergo a greater acoustic attenuation as they propagate through a thermal lesion compared to native tissue.The effect of the increased ultrasound attenuation is expected to be more pronounced in the PA detection geometry of Fig. 3, for which the laser source aperture and the ultrasound transducer are on opposite ends of the sample.Nevertheless, the detected PA signals from HIFU-induced thermal lesions had greater peak-to-peak amplitudes than native tissues.The observed PA contrast between HIFU-treated and native tissues implies that the PA method can indeed detect HIFU-induced thermal lesions at 720 and 845 nm laser illuminations.An understanding of how the detected PA pressure [Eq.( 1)] is influenced by HIFU-induced changes in the optical properties is necessary in order to explain the observed PA contrast in Figs. 6 and 7. respectively, with the SEM for both HIFU-treated and native tissues in the wavelength range of 500-900 nm.The TT and DR results provide a quantitative comparison between HIFU-treated and native tissues in terms of their wavelengthdependent ability to reflect and transmit light between 500 and 900 nm. Figure 9(a) shows that the ratio of TT of native tissue to that of HIFU-treated tissue ranges from 5.54 to 6.66 in the wavelength range of 500-550 nm.However, this ratio tends to gradually decrease as the wavelength increases, reaching a value of 2.74 at 900 nm. Figure 9(b) shows that the ratio of DR of HIFU-treated tissue to that of native tissue increases from about 3.09 at 500 nm to about 3.97 at 900 nm.Figures 9(a) and 9(b) show that the wavelength range of 630-900 nm is appropriate for optical diagnostics of HIFUtreated and native chicken breast tissues in vitro because light has a relatively high transmittance and low reflectance in this range.This observation will be confirmed later with the effective light penetration depth graph.

3.B.2. Optical absorption and reduced scattering coefficients
Using the IAD program, 46 the optical absorption coefficient and the reduced scattering coefficient spectra were determined for each one of the 10 tissue slices.Figures 10(a) and 10(b) show the averaged optical absorption coefficient and the reduced scattering coefficient spectra, respectively, with the SEM for both HIFU-treated and native tissues in the wavelength range of 500-900 nm. Figure 10(a) shows that the optical wavelength range of approximately 670-900 nm has the lowest absorption coefficient for HIFU-treated and native chicken breast tissue, making it an appropriate range for optical diagnostics of relatively deep-seated ROIs.As per Eq. ( 1), the difference in the optical absorption coefficients of HIFU-treated and native tissues in the aforementioned optical wavelength range is a contributor to the difference in the PA response detected from HIFU-treated and native tissues at 720 and 845 nm.In other words, the HIFU- induced increase in the PA signal amplitudes (Figs. 6 and 7) can be attributed, in part, to the increased optical absorption coefficients of HIFU-treated tissues as compared to native tissues.Figure 10(b) shows the large difference between the reduced scattering coefficients of HIFU-treated and native tissues.In fact, at 500 nm, the ratio of the reduced scattering coefficients of HIFU-treated tissue to that of native tissue is ∼11.19, while at 900 nm the ratio increases to about 15.42.The general decrease in the reduced scattering coefficient spectra of both HIFU-treated and native tissues with wavelength, as depicted in Fig. 10(b), is consistent with a previous work that utilized saline-bath heating of tissues. 36lthough they did not use HIFU treatments in their investigation, Yaroslavsky et al. 36 obtained results that are consistent with our result as they have also reported a thermally induced increase in μ a and μ s of saline-bath-treated brain tissues, compared with native brain tissues. 36he observed increase in μ a of HIFU-treated tissues [Fig.10(a)] may be attributable, in part, to the thermally induced formation of methemoglobin, a form of hemoglobin with its iron being in the ferric oxidation state, in the blood. 52he formation of methemoglobin results in an enhanced optical absorption coefficient of the thermally coagulated blood, 52 as compared to native blood, within the wavelength range investigated in our study.Similarly, the increase in μ s of HIFU-treated tissue may be explained by the increase in μ s of thermally coagulated blood compared to native blood. 52t is noteworthy to mention that the visual appearance of the chicken breast tissues (i.e., their pink color) as well as their optical absorption spectrum, which resembles that of blood, 53 attest to the presence of hemoglobin in our samples.A similar argument about the presence of blood residuals in excised tissue samples was made in previous studies that utilized liver 27 and brain 36 tissues.However, due to the low blood content in our tissue samples, the thermally induced formation of methemoglobin might not have had a major contribution to the HIFU-induced changes in the optical properties of the tissue samples.Other factors, discussed in the next paragraph, might have resulted in greater changes in the optical properties.These factors include protein coagulation and dehydration (i.e., water loss).
Protein coagulation and dehydration might have been major reasons behind the changes in the optical properties of HIFU-treated tissues.First, thermal denaturation, which causes functional and structural alterations in the heated proteins, results in the production of highly scattering, amorphous granules to replace well-structured, highly organized molecules, present prior to heating. 54The appearance of the small, thermally denatured granular proteins may be a key factor in the increase in the reduced scattering coefficient of thermally coagulated tissues, based on analyses relying on Mie theory. 55Second, dehydration and tissue shrinkage have been shown to increase the optical absorption coefficient of biologic tissues. 56Moreover, by the virtue of decreasing the overall coagulated tissue volume, dehydration and tissue shrinkage may also result in an increase in the concentration of optical attenuators (scattering and absorbing centres). 36herefore, HIFU-induced dehydration of the treated tissue 053502-10 volume can be an important contributor to the increased optical scattering and absorption of the HIFU-induced thermal lesions in the wavelength range investigated in our experiments.

3.B.3. Effective attenuation coefficient and light penetration depth
Using Eqs. ( 2) and ( 3), the wavelength-dependent effective attenuation coefficients and light penetration depths were determined, respectively, for each one of the 10 chicken breast tissue slices.Figures 11(a Comparing the effective attenuation coefficient and the light penetration depth of HIFU-treated and native tissues can provide an explanation as to why the averaged PA signal amplitudes detected from HIFU-treated tissues were greater than those of native tissues.A superficial HIFU-induced thermal lesion with a lower light penetration depth (i.e., higher effective attenuation coefficient) attenuates greater amount of the incident laser energy than a superficial native tissue of the same volume.In other words, within a HIFU-treated tissue volume, there is a greater probability of photon interactions than a native tissue of the same volume, reducing the light penetration depth and increasing its propagation duration inside the coagulated tissue.Consequently, the amount of the laser energy that is deposited within a HIFU-treated tissue is greater than that within a native tissue, producing a stronger PA pressure rise due to an increase in the laser energy fluence [Eq.( 1 (scattering and absorption) that decrease the effective light penetration depth and confine most of its energy (i.e., increase laser fluence) within the lesion's volume, thereby producing greater PA signal amplitudes than native tissues.

3.C. Monte Carlo simulation of light distribution
The 2D normalized photon fluence within a 5-mm-thick slab of tissue is shown in Figs.12(a) and 12(b) for a chicken breast with and without a HIFU-induced thermal lesion, respectively.Figure 13 shows the unnormalized photon fluence along the axial distance (z) of the tissue at the centerline (i.e., lateral distance = 15 mm) of Fig. 12.Because the tissue optical properties, and thus light distribution, do not dramatically change between 720 and 845 nm, the MC simulation results are shown for the case when the laser illumination is at 720 nm.These results provide a quantitative confirmation to our aforementioned statement about the confinement of laser energy within the volume of the thermal lesion due to the increased light-tissue interactions and the reduced effective light penetration depth (Fig. 11), resulting in an increased photon fluence by ∼2.23-fold in the thermal lesion at and very close (z ≤ 0.8 mm) to the tissue surface, as shown in Fig. 13.We believe that this increase in the photon fluence is a factor contributing to the increase in the PA signal amplitudes of HIFU-treated tissues.
It should be noted that in Fig. 12(a) at axial distance of 3.5 mm, where the boundary between the native and the HIFU-coagulated tissues is located, a vertical discontinuity appears as a result of a sudden change in the scattering coefficient of tissue, causing a sudden change in the mean free path of photons.

CONCLUSIONS
In this study, we have successfully demonstrated the feasibility of PA detection of HIFU-induced thermal lesions in chicken breast tissues in vitro at 720 and 845 nm laser illuminations.A more than threefold increase was observed in the PA signal amplitudes of HIFU-treated tissues compared to native (i.e., untreated) tissues at 720 and 845 nm optical wavelengths, indicating that PA method is indeed capable of detecting HIFU-induced tissue coagulation necrosis and its associated alterations in the molecular and structural compositions of tissues.In order to assist in the interpretation of the aforementioned PA contrast between HIFU-treated and native tissues, we have determined, for the first time, the optical properties of HIFU-induced thermal lesions and compared them with native tissues by performing optical spectroscopy in the wavelength range of 500-900 nm.Based on Eq. ( 1), our spectroscopic investigation has shown that there are direct and indirect optical factors that together contributed to the observed enhancement in the PA pressure as a result of HIFU-induced thermal lesions.The direct optical factor is the increase in the optical absorption coefficient of the HIFUtreated tissue.The indirect optical factor is the large increase in the reduced scattering coefficient, which contributed to an increase in the effective optical attenuation coefficient and a decrease in the light penetration depth.Hence, the increase in the reduced scattering coefficient has indirectly enhanced the PA pressure rise by increasing the absorbed optical energy [Eq.( 1)].For a complete analysis of the effects of the parameters in Eq. ( 1) on the detected PA pressure, assessment of the contribution of thermoacoustic efficiency (Grüneisen parameter) of HIFU-induced thermal lesions on the detected pressure rise should be studied in biologic tissues.Based on the results in Figs.10(a) and 13, from which we determined an approximately 1.13-and 2.23-fold increase in the optical absorption coefficient and photon fluence, respectively, it is expected, based on Eq. ( 1), that the HIFU-induced thermal lesion had a ∼1.46-fold increase in the Grüneisen parameter compared to native tissue, producing a PA signal amplitude increase of about 3.68 at 720 nm.It is also suggested to extend this work to a more clinically relevant scenario by using an in vivo animal model.The blood content in in vivo animal models will have an important role on the generated PA signals from HIFU-treated and native tissues.If methemoglobin is formed in the blood, 30 it is anticipated that the generated PA signals could have greater amplitudes for HIFU-treated tissues compared to native tissues because of the reasons discussed in Sec.3.B.2.On the other hand, if methemoglobin is not formed in the blood, it is expected, as shown in a previous work, 30 that HIFU-induced thermal coagulation could potentially result, depending on the optical illumination wavelength, in a reduction in the PA signal amplitude in regions where there is a large amount of blood, particularly due to reduction of blood plasma and the coagulation of red blood cells in the HIFU-treated region.
The main limitation in this study is the offline detection of HIFU-induced thermal lesions.This was caused by the limited geometry and physical dimensions of the utilized PA imaging system, which did not allow us to combine both HIFU and PA imaging together, although such a combination is possible in principle.For real-time PA imaging and monitoring of HIFU treatments, it is suggested to

FIG. 2 .
FIG. 2. (a) The sequence followed in creating HIFU thermal lesions.(b) A plastic-wrapped, vacuum-sealed chicken breast tissue sample containing HIFU-treated and native ROIs.

FIG. 3 .
FIG. 3. Schematic diagram of the experimental setup of PA detection.In this diagram, the connection between the laser source aperture and the laser assembly does not depict the actual articulated arm for laser beam delivery.
(b), were then attached to a tissue holder and placed Medical Physics, Vol.41, No. 5, May 2014

FIG. 8 .
FIG.8.A comparison between averaged ratios of HIFU-treated and native tissue peak-to-peak PA signal amplitudes from the seven samples obtained at 720 and 845 nm laser illumination.Error bars represent the standard error of the mean (SEM).

Figure 9
Figure9shows the averaged results of the integratingsphere measurements on 10 chicken breast tissue slices.In Figs.9(a) and 9(b), the TT and DR spectra are presented,

FIG. 10 .
FIG. 10. Results of the IAD computations.(a) and (b) show the averaged optical absorption coefficients and reduced scattering coefficients with the SEM, respectively, of HIFU-treated and native chicken breast tissues in the wavelength range of 500-900 nm.
FIG. 11. Results of applying Eqs.(2) and (3).(a) and (b) show the averaged effective attenuation coefficient and effective light penetration depth graphs with the SEM, respectively, of HIFU-treated and native chicken breast tissues in the wavelength range of 500-900 nm.
FIG. 13.Comparison of light fluence as a function of axial distance (depth) in chicken breast tissue with and without a HIFU-induced thermal lesion.The graph represents the unnormalized fluence along the z-axis at the centerline (i.e., lateral distance = 15 mm) of Fig. 12.