Experimental system. A custom dual-linear array transducer (1.54 MHz center frequency) insonified a 25.4-mm cubic tissue-mimicking agarose phantom. Temperature feedback from a needle thermocouple was sent to a PC for closed-loop control of an Antares ultrasound scanner using LabVIEW. The phantom was immersed in a 3-L 32 °C warm water bath, and a 15 °C transducer cooling chamber was placed between the transducer and the warm water bath. The dimensions in this figure are not to scale.
Flow chart for experiments and simulations in this study. LTSLs were synthesized and a fluorescent dye was encapsulated. A tissue-mimicking agarose phantom was created with homogeneously mixed LTSLs. Insonation was applied until thermocouple readings reached desired temperatures of 35, 37, or 39 °C. Fluorescence was quantified by an optical system. FOCUS was then used to simulate the 3D acoustic pressure field, the 3D intensity field was calculated, and an FDTD algorithm was used for simulation of temperature and dye diffusion.
Detailed procedure for calculating dye concentration profile from simulation of temperature and dye diffusion. The temperature profile (T) was calculated using the intensity field of the beam in the second-order FDTD approximation of the bioheat equation. Indexes N and n represent total iteration steps and an individual step, respectively. FI is fluorescence intensity from the dye release and d is the corresponding dye concentration. Conversion between temperature and normalized fluorescence intensity (C1), and conversion between fluorescence intensity and dye concentration (C2), were used to calculate the dye concentration. d accu (n) and D accu (n) represent the accumulated dye concentration profiles at step n before and after the diffusion calculation, respectively. At step n, when the accumulated dye concentration profile after diffusion was calculated (D accu (n)), the difference in dye concentration between steps n and n+1, d diff (n,n+1), was added to D accu (n) and the result was d accu (n+1). The FDTD algorithm was applied on d accu (n+1) to calculate the diffused dye profile, D accu (n+1), and so on. After N iterations, the final accumulated dye concentration profile (D accu (N)) was normalized by the 100%-release dye concentration and compared with experimental results.
Calibration of SIDA fluorescence intensity as a free dye in DPBS. Thirteen serial dilutions from a 0.5 mM solution resulted in a concentration ranging from 0.5 mM (the leftmost point) to 1.22 × 10−4 mM (the rightmost point). The highest fluorescence intensity occurred at a concentration of 6.25 × 10−2 mM (8× dilution of 0.5 mM), indicating a substantial self-quenching effect at the SIDA concentration at or above 6.25 × 10−2 mM. (n = 3; error bars show ± one standard deviation)
Calibration of fluorescence intensity from LTSL release in phantom (a) and from free dye in phantom (b); release efficiency of LTSLs in different solutions (c) and SDS-PAGE analysis of fat-free evaporated milk and milk filtrate (d). (a) LTSL release in the phantom was demonstrated from 23 to 42 °C (solid line, this is also the conversion C1 in Fig. 3 ). The release intensity with 42 °C-treatment was used as 100% release efficiency. There was little release below 33 °C, and the intensity increased dramatically from 35 °C and reached 90% at 37 °C. LTSL release in DPBS showed a rapid increase from 39 °C (dotted line), indicating that the transition temperature of LTSLs in the phantom is lower than that in DPBS. Inset, Zoom-in of the curve between 36 and 42 °C. The release of LTSLs approached ∼100% when the temperature was above 40 °C. A cumulative Gaussian function (dash line) was used to fit the curve by a least-squares method. (b) Correlation between fluorescence intensity of free dye and dye concentration (conversion C2 in Fig. 3 ) in phantom showed high linearity. (c) Release efficiency of LTSLs in different solutions for a comparison of transition temperature. LTSLs in DPBS (dash line), 99% FBS (dotted line), and 99% milk filtrate (black solid line) showed a transition temperature of 39 °C, while LTSLs in the regular evaporated milk (gray solid line) and fat-free evaporated milk (dashed-dotted line) showed a transition temperature of 37 and 32 °C, respectively. (d) MW: Molecular weight ladder; M: fat-free evaporated milk; F: milk filtrate from the fat-free evaporated milk. Bands of the caseins, β-lactoglobulin (β-Lg) and α-lactalbumin (α-La) appeared in the lane M at molecular weight of ∼30, 18, and 14 kDa, respectively. In the lane F, bands of the β-Lg and α-La were distinguishable, and the band corresponding with the caseins was weak. Both lanes M and F were loaded with 30 μg of total protein [In (a)–(c): n = 3; error bars show ± one standard deviation].
Beam profile of single-beam insonation. (a) 3D profile. 3D dimensions of the main lobe were 1.3, 0.9, and 7.0 mm in the elevation, lateral, and depth directions, respectively. The 50%-intensity side lobe had a volume of 0.5% of the main lobe, and the highest intensity of side lobes was 51% of the maximum main lobe intensity. (b) Beam pattern in the depth-elevation plane shows the relative positions of the three lobes, where the contours were the boundaries of 50% intensity. The transducer surface was located at zero on the depth axis.
Optical images and 3D reconstruction of an example case in the 37 °C controlled heating group. (a)–(j) Optical images of phantom slices showed 790 nm fluorescence intensity of SIDA resulting from the release of LTSLs. (a) and (j) were first and last phantom slice with detectable fluorescence, respectively. Each slice had a thickness of 2.0 mm, and spacing between the slices was 2.0 mm. The bar in (a) represents 10 mm in length. (k) and (l) 3D reconstruction of the 50%-release profile. Contribution of the side lobes was observed in the depth-elevation plane (k, dash arrows).
Analysis of 3D release profile in the experiments. Distinct 50%-release spans were shown in the elevation, lateral, and depth directions. Summary of the 50%-release spans were in Table II (n = 5; error bars show ± one standard deviation).
Simulation results and effects of the main lobe and the side lobes. (a) In an example case in the 37 °C controlled heating group, the simulated temperature in the thermocouple tip (dash line) closely matched the thermocouple measurements (solid line) with an average error of 0.04 °C. (b) Gaussian decomposition of the simulated temperature profile (black solid line) in the elevation direction in a case of 37 °C controlled heating. Three Gaussian distributions were initialized in the curve fitting tool, and the results for the main lobe (dash line) and the two side lobes (gray solid lines) were demonstrated. The amplitude of the decomposed Gaussian curves was used as the temperature increase contributed by the main lobe and the side lobes. (c) Temperature increase induced by mainlobe and sidelobes. In (c) n = 5; error bars show ± one standard deviation.
Simulated temperature profiles for 37 °C controlled heating at the thermocouple tip (a) and (b) and dye concentration profiles (c) and (d) in the depth-elevation (a) and (c) and depth-lateral (b) and (d) planes. (a) and (b) In an example case, contours of temperature distribution were marked on both depth-elevation and depth-lateral planes and the highest temperature was 41.8 °C. (c) and (d) In the same case, simulation of normalized dye concentration profile showed a 50% dye release span of 7.7, 6.1, and 17.5 mm in elevation, lateral, and depth directions, respectively.
Overlaid simulated and experimental contours. (a)–(c) The 50% dye-release contours from experiments (black) and 50% normalized dye-concentration contours from simulation (gray) in the depth-elevation plane in the 35, 37, and 39 °C controlled heating groups, respectively. (d)–(f) The 50% contours in the depth-lateral plane in the 35, 37, and 39 °C controlled heating groups, respectively.
Comparison of 50% dye-release spans between experiments and simulation. R2 value of 0.99, 0.95, and 0.80 were calculated in the elevation, lateral, and depth directions, respectively. Solid lines are linear trend lines with intercept set at zero, and dashed lines represent the x = y ideal trend.
Summary of the diluents tested and resulting release temperatures. In each case MPPC or DPPC-containing liposomes (1% of the final volume) were suspended in a phantom or solution with the constituents varied between groups. The groups are ordered from low to high release temperature. The temperature of 10% release is defined as the transition temperature in the remainder of the paper.
Summary of experimental and simulated dye-release spans. (n = 5; error bars show ± one standard deviation).
Article metrics loading...
Full text loading...