Ultrasound molecular imaging detects disease markers on the vascular endothelium by imaging microbubble contrast agents conjugated to targeting antibodies, but standard nonlinear contrast sequences (pulse inversion, contrast pulse sequencing) fail to suppress nonlinear echoes from highly echogenic tissue interfaces, producing false-positive enhancement. Herbst et al. (Investigative Radiology, 2017) showed that interleaving acoustic radiation force (ARF) pulses with pulse inversion, and weighting the images by inter-frame signal decorrelation (ADW-PI), significantly improves contrast-to-tissue ratio in a mouse tumor model — but the physical mechanism of the ARF-induced decorrelation enhancement of adherent microbubbles was not determined: candidate processes (secondary Bjerknes forces, transmit-pressure- or ARF-induced partial detachment, flow shear and erythrocyte collisions) were not separated, the relative contributions of axial phase shifts versus out-of-beam displacement were not quantified, the contrast-to-tissue comparison against clinical contrast pulse imaging rested on commercially processed images of uncertain display mapping, imaging parameters were set empirically, and decorrelation behavior in realistic non-Newtonian low-velocity microvascular flow — including possible microbubble loss over extended imaging — was not characterized.

Background

Ultrasound molecular imaging aims to visualize disease markers on the vascular endothelium by injecting microbubble contrast agents — gas-filled bubbles of roughly 1–10 μm with lipid or protein shells — that are conjugated to targeting antibodies, for example against vascular endothelial growth factor receptor 2 (VEGFR2), a marker of tumor angiogenesis. Because microbubbles oscillate nonlinearly under ultrasound insonation, contrast-specific sequences can suppress tissue signal: pulse inversion (PI) transmits two phase-inverted pulses and sums the echoes, so that linear tissue reflections cancel while nonlinear bubble echoes survive; contrast pulse sequencing (CPS) is a multi-pulse amplitude- and phase-modulated scheme built on the same principle. Both can nevertheless fail at highly echogenic tissue interfaces such as bone or the skin line, which generate nonlinear reflections of their own and cause false-positive enhancement. A standard figure of merit is the contrast-to-tissue ratio (CTR): the ratio of microbubble signal inside a region of interest to the signal from background tissue.

A separate line of work exploits microbubble motion rather than nonlinearity. Inter-frame signal decorrelation — the frame-to-frame loss of correlation of the echo signal — is elevated for microbubbles relative to static tissue. Wang et al. (2013) showed that in large vessels this statistical signature of adherent (bound) bubbles is caused by shear forces from blood flow, which drive continuous aggregation, dislodgement, and out-of-beam motion, and that the signature can be measured by statistics such as the normalized singular spectrum area (NSSA), a parameter derived from the singular values of the echo-data matrix that is monotonic in signal decorrelation. The same group (Wang et al., 2015) optically verified that acoustic radiation force (ARF) — the time-averaged force an acoustic wave exerts on a scatterer, which can translate microbubbles without destroying them — redistributes and displaces microbubbles in large vessels.

Herbst et al. (2017) combined these ideas into ARF decorrelation-weighted pulse inversion (ADW-PI): long ARF pulses interleaved with PI imaging in a mouse hindlimb tumor model (MC38, a murine colon carcinoma cell line, tumors grown in C57BL/6 mice, anti-VEGFR2 lipid-shelled microbubbles, a Verasonics programmable scanner, with a Siemens Sequoia clinical scanner providing CPS comparison images). They found that microbubble-bound tumor regions decorrelated significantly more than static tissue, that applying ARF significantly increased microbubble decorrelation without bubble destruction, and that weighting PI images by the measured decorrelation improved CTR by approximately 3 dB over PI and 2 dB over CPS. The mechanism of the ARF-induced decorrelation enhancement, however, was left undetermined. Three candidate processes were discussed. First, secondary Bjerknes forces — time-averaged mutual forces between bubbles driven by one another's scattered fields — which Kokhuis et al. (2013) showed can rapidly deform targeted microbubbles into prolate shapes and attract them toward one another. Second, partial detachment of bound bubbles: the study's transmit pressures (286 kPa peak negative for PI, 255 kPa for ARF) exceeded the roughly 150 kPa level at which Loughran et al. (2012) found secondary radiation forces can overcome the molecular binding of adherent microbubbles, so some detachment may occur even without ARF, with ARF plausibly increasing it. Third, shear forces and erythrocyte collisions in the tumor microvasculature: Owen et al. (2014) showed in 200-μm tubes that collisions with red blood cells perturb and dislodge targeted bubbles. Crucially, the ARF-versus-no-ARF decorrelation difference was statistically significant only when decorrelation was computed from complex in-phase/quadrature (I/Q) echo data, which retains phase, and vanished with magnitude-only data — implying that axial microbubble motion (which shifts echo phase) contributes — but the relative contributions of axial phase shifts and out-of-beam (lateral/elevational) displacement were never quantified. In addition, the CTR comparison with CPS relied on commercially post-processed images whose display mapping was assumed linear, and the ARF transmit parameters (frequency, pulse duration, peak negative pressure and mechanical index — the regulatory safety index MI, defined as peak negative pressure divided by the square root of acoustic frequency in MHz), the parameters of the Gaussian remapping filter (which maps measured decorrelation values onto image weights through a Gaussian-shaped function), and the display weighting parameter α (an exponent controlling how strongly the decorrelation weight is applied to the displayed image) were all set by empirical optimization rather than systematic study.

Problem Statement

The problem left open by Herbst et al. (2017) is to establish the physical basis of acoustic-radiation-force-induced signal decorrelation as a microbubble-specific contrast mechanism for non-destructive ultrasound molecular imaging, and to validate its contrast performance quantitatively. Its components, which support this single objective, are:

  1. Mechanism identification and quantification. Determine which physical processes produce the measured ARF enhancement of inter-frame decorrelation of adherent, molecularly targeted microbubbles in vivo — (a) secondary Bjerknes (secondary radiation) forces deforming bound bubbles into prolate shapes or migrating them laterally, (b) partial detachment of bound bubbles induced by ARF or by the imaging transmit pressures themselves, and (c) shear forces and erythrocyte collisions in the tumor microvasculature — and quantify each mechanism's contribution, reporting the relative contributions as explicit quantities together with the assumptions and evidence on which they rest. In particular, decompose the enhancement into the contribution of axial microbubble motion, which appears as phase shifts in complex I/Q echo data (the enhancement is statistically significant only when complex echo data are used), versus out-of-beam (lateral/elevational) displacement; the source study observed this distinction but did not quantify it.

  2. Quantitative benchmarking. Quantify the contrast-to-tissue ratio of decorrelation-weighted pulse inversion against contrast pulse sequencing on raw, unprocessed echo data or under standardized linear display mappings, removing the confound of unknown nonlinear processing in commercially displayed CPS images. State the evidence and metric on which the comparison rests and, for any quantity the available data cannot supply, specify and justify the measurement that would produce it.

  3. Realism and durability. Characterize the decorrelation behavior of adherent microbubbles in realistic microvascular flow conditions (non-Newtonian rheology, low flow velocity), and quantify potential microbubble loss — detachment or destruction — over extended imaging intervals, since sustained non-destructive imaging is required for clinical translation. This component addresses a flow regime distinct from the in vivo setting of component 1 and must likewise state the evidence it rests on and specify and justify any measurement required.

A complete answer addresses all three components above: it provides a mechanistically grounded, quantitatively validated account of the ARF enhancement of inter-frame decorrelation of adherent targeted microbubbles — identifying which physical processes produce the enhancement and their relative contributions, including the split between axial phase shifts and out-of-beam displacement; it gives an independent quantitative comparison of the contrast-to-tissue ratio of decorrelation-weighted pulse inversion against contrast pulse sequencing, rather than restating the values already reported in the Background; and it characterizes decorrelation and microbubble loss in a realistic microvascular-flow regime distinct from the in vivo setting of the first component — so that a practitioner can judge whether ARF-decorrelation imaging is a reliable microbubble-specific contrast mechanism for molecular imaging of angiogenesis and related endothelial markers.

References

  1. Herbst EB, Unnikrishnan S, Wang S, Klibanov AL, Hossack JA, Mauldin FW. The Use of Acoustic Radiation Force Decorrelation-Weighted Pulse Inversion for Enhanced Ultrasound Contrast Imaging. Investigative Radiology 52(2):95-102, 2017. DOI: 10.1097/RLI.0000000000000313. https://doi.org/10.1097/RLI.0000000000000313
  2. Wang S, Mauldin FW, Klibanov AL, Hossack JA. Shear Forces from Flow Are Responsible for a Distinct Statistical Signature of Adherent Microbubbles in Large Vessels. Molecular Imaging 12(6), 2013. DOI: 10.2310/7290.2013.00057. https://doi.org/10.2310/7290.2013.00057
  3. Wang S, Wang CY, Unnikrishnan S, Klibanov AL, Hossack JA, Mauldin FW. Optical Verification of Microbubble Response to Acoustic Radiation Force in Large Vessels With In Vivo Results. Investigative Radiology 50(11):772-784, 2015. DOI: 10.1097/RLI.0000000000000185. https://doi.org/10.1097/RLI.0000000000000185
  4. Kokhuis TJA, Garbin V, Kooiman K, Naaijkens BA, Juffermans LJM, Kamp O, van der Steen AFW, Versluis M, de Jong N. Secondary Bjerknes Forces Deform Targeted Microbubbles. Ultrasound in Medicine & Biology 39(3):490-506, 2013. DOI: 10.1016/j.ultrasmedbio.2012.09.025. https://doi.org/10.1016/j.ultrasmedbio.2012.09.025
  5. Loughran J, Sennoga C, Eckersley RJ, Tang M-X. Effect of ultrasound on adherent microbubble contrast agents. Physics in Medicine and Biology 57(21):6999-7014, 2012. DOI: 10.1088/0031-9155/57/21/6999. https://doi.org/10.1088/0031-9155/57/21/6999
  6. Owen J, Grove P, Rademeyer P, Stride E. The influence of blood on targeted microbubbles. Journal of the Royal Society Interface 11(100):20140622, 2014. DOI: 10.1098/rsif.2014.0622. https://doi.org/10.1098/rsif.2014.0622
  7. Herbst EB, Unnikrishnan S, Klibanov AL, Mauldin FW, Hossack JA. Validation of Normalized Singular Spectrum Area as a Classifier for Molecularly Targeted Microbubble Adherence. Ultrasound in Medicine & Biology 45(9):2493-2501, 2019. DOI: 10.1016/j.ultrasmedbio.2019.05.026. https://doi.org/10.1016/j.ultrasmedbio.2019.05.026
  8. Herbst EB, Klibanov AL, Hossack JA, Mauldin FW. Dynamic Filtering of Adherent and Non-adherent Microbubble Signals Using Singular Value Thresholding and Normalized Singular Spectrum Area Techniques. Ultrasound in Medicine & Biology 47(11):3240-3252, 2021. DOI: 10.1016/j.ultrasmedbio.2021.06.019. https://doi.org/10.1016/j.ultrasmedbio.2021.06.019
  9. Zhao F, Unnikrishnan S, Herbst EB, Klibanov AL, Mauldin FW, Hossack JA. A Targeted Molecular Localization Imaging Method Applied to Tumor Microvasculature. Investigative Radiology 56(4):197-206, 2021 (online first September 24, 2020). DOI: 10.1097/RLI.0000000000000728. https://doi.org/10.1097/RLI.0000000000000728
  10. Huang Y, Herbst E, Hossack JA. Simulation of microbubble displacement and normalized singular spectrum area measurements of microbubble signals in a vessel phantom. Proceedings of the 2021 IEEE International Ultrasonics Symposium (IUS), pp. 1-3, 2021. DOI: 10.1109/IUS52206.2021.9593771. https://doi.org/10.1109/IUS52206.2021.9593771
  11. Huang Y, Herbst EB, Xie Y, Yin L, Islam ZH, Kent EW, Wang B, Klibanov AL, Hossack JA. In Vivo Validation of Modulated Acoustic Radiation Force-Based Imaging in Murine Model of Abdominal Aortic Aneurysm Using VEGFR-2-Targeted Microbubbles. Investigative Radiology 58(12):865-873, 2023. DOI: 10.1097/RLI.0000000000001000. https://doi.org/10.1097/RLI.0000000000001000
  12. Huang Y, Zhao F, Xie Y, Mauldin FW, Klibanov AL, Hossack JA. High-Resolution Ultrasound Molecular Imaging With Incremental Burst Sequence: In Vitro and in Vivo Validation. IEEE Transactions on Ultrasonics, Ferroelectrics, and Frequency Control 73(4):445-459, 2026 (online first June 2025). DOI: 10.1109/TUFFC.2025.3576337. https://doi.org/10.1109/TUFFC.2025.3576337