Calibrated noninvasive temperature mapping from the change in backscattered ultrasound energy (CBE), including the second-harmonic variant introduced to reduce motion sensitivity, was first demonstrated quantitatively in ex vivo pork muscle in a single tissue/transducer configuration with offline processing (Shaswary et al., Ultrasonics 2021). Follow-up work by the same group has since established real-time closed-loop control of mild hyperthermia with high-frequency CBE thermometry in ex vivo tissue and in the hind leg of anesthetized mice (about +/-1 degree C versus thermocouple). What remains open is the generalization implied by the ex vivo scope demonstrated by Shaswary et al. (2021): a calibration that transfers across tissue types, imaging parameter sets, and individual subjects in heterogeneous and perfused tissue; robustness to physiologic motion and treatment-induced tissue displacement; and in vivo validation beyond a single homogeneous muscle site, including of the harmonic-CBE variant itself, with temperature-triggered thermosensitive-liposome drug release as the motivating end application.

Background

Mild hyperthermia treats tissue at approximately 41-45 degrees C, a narrow window used, for example, to trigger drug release from thermosensitive liposomes while avoiding thermal damage; holding tissue in this window requires temperature feedback during heating. Magnetic resonance thermometry is the only clinically validated noninvasive feedback but is expensive and comparatively slow, which motivates ultrasound-based thermometry: ultrasound is portable, non-ionizing, and acquires data fast. Three families of ultrasound thermometry exist. The echo-shift (time-shift) method infers temperature from shifts of radio-frequency (RF) echoes caused by temperature-dependent sound speed and thermal expansion; it produces 2D maps but is strongly degraded by tissue motion. The change in backscattered energy (CBE) method exploits the temperature dependence of scattered ultrasound power, first analyzed theoretically by Straube and Arthur (1994); CBE is typically computed region-wise in decibels as the ratio of backscattered power at the current temperature to that at a baseline. Finally, nonlinear propagation of an ultrasound pulse in tissue generates acoustic harmonics (integer multiples of the transmit frequency) whose amplitudes are temperature dependent. Maraghechi, Kolios and Tavakkoli (2019) showed that the CBE of acoustic harmonics can localize a heated region in phantoms and ex vivo bovine muscle, but without calibration to absolute temperature. Shaswary et al. (2021) closed that gap ex vivo: using low-intensity focused ultrasound (LIFU) heating of pork muscle over 37-47 degrees C, they calibrated the second-harmonic CBE (CBEh2) and a combined fundamental-plus-second-harmonic CBE (CBEcomb) against needle thermocouples, produced 2D temperature maps from RF data, compensated translational axial and lateral motion, and demonstrated closed-loop temperature control with pre-determined heating parameters. A companion method, nonlinear frequency compounding, improved the signal-to-noise ratio of harmonic-CBE maps ex vivo (Hornsby, Shaswary and Tavakkoli, 2021). The remaining agenda implied by that ex vivo demonstration: calibration must be repeated for each tissue type and imaging parameter set and inter-subject variability in heterogeneous tissue requires larger calibration studies; processing was offline, with real-time frame rates (tens of frames per minute) needing parallel computing; motion compensation addressed only translational displacements, not the physiologic motion and treatment-induced tissue displacement encountered in vivo; the closed-loop experiment lacked simultaneous thermocouple validation; and extension to in vivo models and other heating modalities (laser, alternating magnetic field) was unrealized, with combined operation with thermosensitive-liposome drug delivery remaining a motivating end application rather than a demonstrated capability. Subsequent work by the same group advanced the CBE family: a 2021 study presented at the IEEE International Ultrasonics Symposium (IUS) demonstrated real-time closed-loop control with CBE from a clinical high-frequency scanner during interstitial laser heating of ex vivo porcine tissue (37 to 43 degrees C, about +/-0.8 degrees C versus fiber-optic reference), and Nguyen et al. (2024) demonstrated real-time proportional-integral-derivative closed-loop control with high-frequency CBE thermometry both ex vivo and in the hind leg of anesthetized mice, holding the focus near 40 degrees C for about 4 minutes with in vivo accuracy of about +/-0.97 degrees C versus a calibrated thermocouple, describing the result as a foundation for an ultrasound-mediated drug-delivery system. Adjacent results do not resolve the open problem: single-phase CBE imaging has been validated clinically only for monitoring the gas-bubble ablation zone during radiofrequency ablation of liver tumors (Wang et al., 2022), not for calibrated hyperthermia-range temperature; CBE-based thermometry has been extended to ablative temperatures (up to about 90 degrees C) only ex vivo (Rohfritsch et al., 2024); and data-driven CBE approaches from B-mode images remain at phantom or ex vivo scope (Oliveira et al., 2024). Reviews of quantitative-ultrasound thermometry (Li et al., 2022; Wang et al., 2025) consistently identify calibration dependence on tissue type and equipment, motion sensitivity, and the absence of in vivo and clinical validation as the field's unresolved limitations.

Problem Statement

Can calibrated noninvasive ultrasound thermometry based on backscattered energy (including the second-harmonic CBE variant introduced by Shaswary et al., 2021) be generalized from a single ex vivo tissue and transducer configuration to in vivo heterogeneous, perfused, moving tissue, so that it provides clinically usable, real-time temperature feedback for closed-loop mild hyperthermia? The task is to determine whether and how this generalization is achievable and to deliver a concrete method and validation plan for it; the open problem comprises the following facets, which are the aspects that determination must address rather than four independently required experimental demonstrations: (a) determine how the CBE-to-temperature calibration depends on tissue type, tissue heterogeneity, perfusion, and imaging parameters, and establish a calibration that transfers across tissue types, parameter sets, and individual subjects, or a self-calibrating scheme that removes the need for per-session recalibration; (b) achieve robustness to physiologic motion, extending motion compensation beyond the translational motion correction demonstrated ex vivo to the physiologic motion and treatment-induced tissue displacement encountered in vivo; (c) validate the calibrated estimates in vivo with simultaneous invasive thermometry in heterogeneous, perfused tissue across multiple tissue types and subjects, at accuracy on the order of 1 degree C within the 41-45 degrees C mild-hyperthermia window and at frame rates sufficient for real-time closed-loop feedback; and (d) demonstrate the harmonic-CBE variant itself in vivo, since the existing in vivo demonstration used high-frequency CBE. Temperature-triggered drug-delivery applications, such as release from thermosensitive liposomes, which require holding tissue at 41-45 degrees C under ultrasound temperature feedback, are the motivating end-application context of this goal rather than a required completion element.

References

  1. Shaswary E, Assi H, Yang C, Kumaradas JC, Kolios MC, Peyman G, Tavakkoli J (2021). Noninvasive calibrated tissue temperature estimation using backscattered energy of acoustic harmonics. Ultrasonics 114:106406. DOI: 10.1016/j.ultras.2021.106406. https://doi.org/10.1016/j.ultras.2021.106406
  2. Maraghechi B, Kolios MC, Tavakkoli J (2019). Feasibility of detecting change in backscattered energy of acoustic harmonics in locally heated tissues. International Journal of Hyperthermia 36(1):963-973. DOI: 10.1080/02656736.2019.1660001. https://doi.org/10.1080/02656736.2019.1660001
  3. Hornsby T, Shaswary E, Tavakkoli J (2021). Development of an ultrasonic nonlinear frequency compounding method with applications in tissue thermometry. Journal of the Acoustical Society of America 150(4):3192-3203. DOI: 10.1121/10.0006207. https://doi.org/10.1121/10.0006207
  4. Shaswary E, Assi H, Yang C, Kumaradas JC, Kolios MC, Peyman G, Tavakkoli J (2021). Real-time non-invasive control of tissue temperature using high-frequency ultrasonic backscattered energy. Proceedings of the 2021 IEEE International Ultrasonics Symposium (IUS):1-4. DOI: 10.1109/IUS52206.2021.9593438. https://doi.org/10.1109/IUS52206.2021.9593438
  5. Nguyen M, Agarwal A, Kumaradas JC, Kolios MC, Peyman G, Tavakkoli J (2024). Real-time non-invasive control of ultrasound hyperthermia using high-frequency ultrasonic backscattered energy in ex vivo tissue and in vivo animal studies. Physics in Medicine and Biology 69(21):215001. DOI: 10.1088/1361-6560/ad7f19. https://doi.org/10.1088/1361-6560/ad7f19
  6. Li S, Zhou Z, Wu S, Wu W (2022). A Review of Quantitative Ultrasound-Based Approaches to Thermometry and Ablation Zone Identification Over the Past Decade. Ultrasonic Imaging 44(4):213-228. DOI: 10.1177/01617346221120069. https://doi.org/10.1177/01617346221120069
  7. Wang X, Zhao S, Zhang A (2025). Image-Based Monitoring of Thermal Ablation. Bioengineering 12(1):78. DOI: 10.3390/bioengineering12010078. https://doi.org/10.3390/bioengineering12010078
  8. Wang CY, Zhou Z, Chang YH, Ho MC, Lu CM, Wu CH, Tsui PH (2022). Ultrasound single-phase CBE imaging for monitoring radiofrequency ablation of the liver tumor: A preliminary clinical validation. Frontiers in Oncology 12:894246. DOI: 10.3389/fonc.2022.894246. https://doi.org/10.3389/fonc.2022.894246
  9. Rohfritsch A, Barrere V, Estienne L, Melodelima D (2024). 2D ultrasound thermometry during thermal ablation with high-intensity focused ultrasound. Ultrasonics 142:107372. DOI: 10.1016/j.ultras.2024.107372. https://doi.org/10.1016/j.ultras.2024.107372
  10. Oliveira LFR, Franca FMG, Pereira WCA (2024). A Data-Driven Approach for Estimating Temperature Variations Based on B-mode Ultrasound Images and Changes in Backscattered Energy. Ultrasonic Imaging 46(1):3-16. DOI: 10.1177/01617346231205810. https://doi.org/10.1177/01617346231205810
  11. Straube WL, Arthur RM (1994). Theoretical estimation of the temperature dependence of backscattered ultrasonic power for noninvasive thermometry. Ultrasound in Medicine and Biology 20(9):915-922. DOI: 10.1016/0301-5629(94)90051-5. https://doi.org/10.1016/0301-5629(94)90051-5