The 2021–2023 simulation studies of Wang and co-workers proposed "thermostatic" HIFU hyperthermia in which a 256-element ring ultrasound array heats a thigh tumor while full-waveform-inversion speed-of-sound (SOS) imaging, converted to temperature through a linear SOS–temperature calibration, closes the feedback loop; in k-Wave simulations the tumor held 42–43 °C within ±0.2 °C for 5 hours, but under strong simplifications (fixed bone/vessel SOS, exact B-mode segmentation, no noise, no physiological motion or variability). The authors' stated open agenda is to combine the model with actual experiments — the 2023 journal version specifies producing the ring array transducer as designed and gradually testing it on phantom and experimental animals to verify the reliability of the thermal therapy model — and to extend the treatment from the thigh to other parts of the body. Companion ring-array ultrasound-tomography work has now demonstrated SOS-based temperature monitoring ex vivo and in vitro (Pattyn et al. 2022; Bustamante et al. 2024, with 0.58 °C maximum thermometry error in vitro), and the source group's 2026 preclinical prototype achieved reliable SOS temperature monitoring during ex vivo hyperthermia heating with a size-adjustable ring array. What remains open is the actual claim of the simulation: closed-loop thermostatic control driven by SOS-image feedback in living, perfused tissue — multi-hour sessions with quantified regulation accuracy under real physiological conditions (tissue-dependent and sign-reversed dSOS/dT, vessel motion, imperfect segmentation, measurement noise) — and extension beyond thigh geometry.

Background

Therapeutic mild hyperthermia aims to hold tumor tissue at approximately 42–43 °C for an hour or more, a range in which heating sensitizes tumors to radiotherapy and chemotherapy without destroying tissue outright. Its clinical success hinges on temperature regulation: the tumor must stay inside the therapeutic window for the whole session while surrounding healthy tissue is spared, so the therapy needs a non-invasive thermometer and a controllable heat source acting together in a closed loop. Magnetic-resonance-guided focused ultrasound, whose proton-resonance-frequency thermometry provides real-time temperature maps, is the reference platform but requires large and expensive scanners; purely ultrasound-guided systems are cheap and portable but lack a non-invasive temperature readout.

The candidate source (Wang, Wang, Yuan and Carson, arXiv:2109.05709, 2021; journal version Wang et al., Applied Sciences, 2023) proposes an "all-acoustic" alternative demonstrated entirely in simulation. A 256-element ring ultrasound transducer array surrounds a human thigh model in a 30 °C water bath and performs three jobs with one aperture: B-mode (pulse-echo) imaging recovers tissue boundaries; high-intensity focused ultrasound (HIFU) electronically focuses heating into the tumor; and speed-of-sound (SOS) imaging reconstructs the sound-speed distribution of the cross-section, from which temperature is read. The physical basis is the temperature dependence of sound speed: in water-like and muscle-like tissues it rises by roughly +1 m/s per °C near body temperature (in fat it decreases), so a local change in the reconstructed SOS map converts into a local temperature change. Reconstruction uses full-waveform inversion (FWI), which iteratively adjusts a sound-speed model until waveforms simulated by the acoustic wave equation (here with the open-source k-Wave pseudospectral solver) match the transmissions recorded at opposing array elements; unlike straight-ray filtered back-projection or bent-ray time-of-flight tomography, FWI accounts for diffraction and scattering. Heating follows the Pennes bioheat equation, the standard model combining thermal diffusion, heat loss to blood perfusion, and acoustic heat deposition, and a period-based feedback controller rescales the heat source every 60 s from the temperature change of the previous period.

In three-dimensional thigh simulations with 42 °C and 43 °C targets, this controller held the tumor-center temperature within ±0.2 °C for 5 hours, and a tumor-region SOS fluctuation below 0.2 m/s was mapped to 0.2 °C monitoring accuracy. These numbers rest on strong simplifications, all acknowledged in the paper: the SOS of bone, bone marrow and blood vessels is fixed at room-temperature standard values; the temperature–SOS relation is taken as a direct proportion applied uniformly; the FWI is initialized from exact B-mode segmentation of seven tissue compartments, with small-vessel regions excluded from the inversion because their SOS iteration is noise-sensitive; and no measurement noise, physiological motion, or heterogeneous perfusion is present. The authors' conclusion states the remaining agenda plainly: the model must be combined with actual experiments, and only then can treatment extend to other parts of the body.

A companion experimental line exists for the same hardware concept (the 2021 paper notes its ring array matches equipment at Wayne State University). Pattyn et al. (2022) showed ex vivo that ring-array ultrasound tomography tracks temperature through sound-speed changes in tissue phantoms, but with a systematic bias of their ray-based inversion, which they propose to fix with waveform-based inversion. Bustamante et al. (2024) showed in vitro that a ring array can generate and maintain mild hyperthermia in a tissue-mimicking phantom, and that ultrasound-tomography thermometry tracks a cooling water inclusion with 0.58 °C maximum error — while explicitly framing the fully closed, image-guided all-acoustic system as the still-unrealized goal. In parallel, reflection-mode SOS-shift thermometry (Grutman and Ilovitsh, 2023) reconstructs temperature maps from B-mode pixel displacements with errors below 2 °C for changes of 8 °C and larger in ex vivo tissue — valuable for ablation monitoring but not yet at the precision closed-loop thermostatic control demands. No study has yet demonstrated the actual claim of the simulation: multi-hour, closed-loop thermostatic control whose feedback comes from SOS images, working in real perfused tissue.

Problem Statement

Determine, by experiment, whether the SOS-imaging-guided thermostatic hyperthermia demonstrated in simulation works in real tissue — physically closing the loop and quantifying what survives when the simplifying assumptions are relaxed. The common objective joining all components is a single integrated ultrasound system that heats and measures with one ring array, with the controller's feedback taken from reconstructed speed-of-sound images rather than interstitial sensors. Concretely, an answer must establish:

  1. Closed-loop control from SOS feedback in real tissue: an integrated ring-array system performs multi-hour mild hyperthermia (targets on the order of 42–43 °C) while its feedback signal is the SOS-derived temperature, with performance judged against independent ground-truth thermometry (implanted thermocouples or fiberoptic probes, or MR thermometry), quantifying regulation accuracy relative to the ±0.2 °C level achieved in simulation — or documenting the degradation.

  2. Anatomical generality: whether performance survives geometries other than the thigh model — larger bones, air interfaces, differing perfusion rates (for example the breast, as the journal version specifies).

The completion boundary follows the journal version's stated experimental agenda — "produce the ring array transducer as designed and gradually test it on phantom, experimental animals to verify the reliability of our proposed thermal therapy model": an in vivo (experimental animal, or human) demonstration of SOS-feedback thermostatic hyperthermia with quantified regulation accuracy under physiological conditions, with the same in vivo demonstration also required in at least one non-thigh geometry (for example the breast); phantom and ex vivo demonstrations are partial progress on this graduated agenda, not completion.

References

  1. Yuchan Wang, Yuening Wang, Jie Yuan, Paul Carson, 'Thermostatic Hyperthermia with Non-invasive Temperature Monitoring through Speed of Sound Imaging', arXiv:2109.05709 (2021). https://arxiv.org/abs/2109.05709
  2. Haoyang Wang, Yuchen Sun, Yuxin Wang, Ying Chen, Yun Ge, Jie Yuan, Paul Carson, 'Temperature-Controlled Hyperthermia with Non-Invasive Temperature Monitoring through Speed of Sound Imaging', Applied Sciences 13(12):7317 (2023). DOI: 10.3390/app13127317. https://doi.org/10.3390/app13127317
  3. A. Pattyn, K. Kratkiewicz, N. Alijabbari, P. L. Carson, P. Littrup, J. B. Fowlkes, N. Duric, M. Mehrmohammadi, 'Feasibility of ultrasound tomography-guided localized mild hyperthermia using a ring transducer: Ex vivo and in silico studies', Medical Physics 49(9):6120–6136 (2022). DOI: 10.1002/mp.15829. https://doi.org/10.1002/mp.15829
  4. David Bustamante, Yan Yan, Trevor Mitcham, Rehman Ali, Brian Marples, Kimberly R. Gergelis, Peter Littrup, Nebojsa Duric, Mohammad Mehrmohammadi, 'Generating and monitoring mild hyperthermia using a ring array ultrasound transducer', International Journal of Hyperthermia 41(1):2376681 (2024). DOI: 10.1080/02656736.2024.2376681. https://doi.org/10.1080/02656736.2024.2376681
  5. David Bustamante, Yan Yan, Nebojsa Duric, Mohammad Mehrmohammadi, 'Acoustic Waveform Optimization for Improved Mild Hyperthermia in Cancer Therapy', 2024 IEEE Ultrasonics, Ferroelectrics, and Frequency Control Joint Symposium (UFFC-JS), pp. 1–4 (2024). DOI: 10.1109/UFFC-JS60046.2024.10793660. https://doi.org/10.1109/UFFC-JS60046.2024.10793660
  6. David Bustamante, Yan Yan, Trevor Mitcham, Nebojsa Duric, Mohammad Mehrmohammadi, 'The efficacy of an ultrasound ring-array transducer for hyperthermia generation', Proceedings of SPIE, Medical Imaging 2024: Ultrasonic Imaging and Tomography (2024). DOI: 10.1117/12.3006248. https://doi.org/10.1117/12.3006248
  7. Ze Li, Yuchen Sun, Lei Li, Zhendong Yao, Xiao Yin, Jie Yuan, 'Radius-Adjustable Ring Ultrasound Transducer Array for Real-Time 3D Imaging of Large Volume Targets', Ultrasound in Medicine & Biology 51:2161–2172 (2025). DOI: 10.1016/j.ultrasmedbio.2025.08.015. https://doi.org/10.1016/j.ultrasmedbio.2025.08.015
  8. Yuchen Sun, Ze Li, Lei Li, Jie Yuan, Nishan Dong, Junjie Lu, 'Noninvasive Temperature Monitoring Hyperthermia With Size-Adjustable Ring Array', IEEE Transactions on Instrumentation and Measurement 75:1–10 (2026). DOI: 10.1109/TIM.2026.3652693. https://doi.org/10.1109/TIM.2026.3652693
  9. Tal Grutman, Tali Ilovitsh, 'Dense speed-of-sound shift imaging for ultrasonic thermometry', Physics in Medicine & Biology 68(21):215004 (2023). DOI: 10.1088/1361-6560/acfec3. https://doi.org/10.1088/1361-6560/acfec3
  10. Gaya Lamm, Tal Grutman, Mike Bismuth, Tali Ilovitsh, 'Cryoablation temperature monitoring with dense ultrasonic speed-of-sound shift imaging', bioRxiv preprint (2026). DOI: 10.64898/2026.01.04.697528. https://doi.org/10.64898/2026.01.04.697528