During HIFU ablation of ex vivo porcine fat, the X-ray CT Hounsfield unit (HU) at the focus follows a quadratic trajectory in temperature with a minimum (nadir) near 44.5 degrees Celsius marking hyperthermic injury onset, and the heating/cooling HU-temperature hysteresis correlates with the maximal temperature reached (Weiss, Sosna, Goldberg, Azhari, 2014). Whether these image-only temperature and damage indicators survive in vivo perfusion, clinically acceptable CT radiation dose, and validation against biological damage metrics remains open, with the motivating clinical application being HIFU of tumours surrounded by fatty tissue such as breast tumours: a 2023 in vivo CT thermography study covered only perfused porcine liver with a linear-regime ablation-zone approach, and the fat-specific nadir/hysteresis framework has never been tested in vivo. The problem is to determine, by controlled in vivo experiment, the fate of the quadratic trajectory, the nadir, and the hysteresis indices in perfused fatty tissue, the minimum clinically acceptable dose and protocol that preserve their fidelity, and whether the indices predict biological damage.

Background

High-intensity focused ultrasound (HIFU) is a non-invasive thermal therapy in which an ultrasound beam is concentrated at a focal zone deep inside tissue to raise it to cytotoxic temperature; it is applied clinically to uterine fibroids, prostate tissue, bone metastases, and neurological targets, and fat-containing organs such as the breast are a long-standing target of interest. Effective HIFU needs non-invasive real-time temperature monitoring: underheating leaves tumor viable, overheating destroys adjacent healthy tissue. The only clinically validated non-invasive thermometry for HIFU is magnetic-resonance thermometry based on the proton resonance frequency (PRF) shift — the temperature-dependent shift of the water resonance in MR images — which is accurate in water-like tissue but requires costly MRI-compatible equipment, limits patient access and throughput, and degrades in fat, where fat-suppression techniques are needed. X-ray computed tomography (CT) offers a potential alternative: the CT attenuation value, quoted in Hounsfield units (HU), depends on physical density, which changes with temperature; in water-like soft tissue the HU falls roughly linearly with heating (about -0.4 to -0.6 HU per degree Celsius in ex vivo liver). Weiss, Sosna, Goldberg, and Azhari (2014) showed that fatty tissue behaves strikingly differently during HIFU heating: in six ex vivo porcine subcutaneous fat specimens, the mean HU at the focus followed a parabolic (quadratic) trajectory versus temperature (R2>0.87R^2 > 0.87, where R2R^2 denotes the coefficient of determination of the fit), first decreasing to a minimum — the nadir — at a mean temperature of about 44.5 ∘^\circC, near the onset of hyperthermic injury in fat, and then increasing as heating progressed; during post-ablation cooling the HU returned along an approximately linear trajectory (R2≥0.9R^2 \ge 0.9). Their companion phantom and ex vivo study (Weiss et al. 2014) showed that this heating/cooling mismatch is a HU-temperature hysteresis — a loop traced by the HU-temperature curve when heating and cooling follow different paths — whose magnitude correlates with the maximal temperature reached (R2>0.9R^2 > 0.9 in albumin phantoms), suggesting quantitative damage indices such as loop indentation-depth measures. Because the nadir appears as a minimum on the HU-versus-time curve, injury onset can in principle be flagged from images alone, without absolute temperature calibration. These results were obtained exclusively ex vivo, on excised non-perfused porcine fat, with no radiation-dose constraint. Whether the framework survives in vivo — where blood perfusion, physiological hydration, motion, and tissue heterogeneity reshape both the temperature field and the HU trajectories — and whether it does so at a clinically acceptable CT dose, was left explicitly open and defines the translational question.

Problem Statement

Determine whether the ex vivo CT HU monitoring framework for HIFU in fatty tissue — the quadratic heating-stage HU-temperature trajectory, the nadir (the minimum of the HU-temperature curve) near 44.5 ∘^\circC marking hyperthermic injury onset, and the hysteresis indices (quantitative measures of the loop traced by the HU-temperature curve between the heating and cooling stages) that correlate with the maximal temperature reached — remains valid in vivo, and establish the conditions under which it could be used clinically. The single overarching objective is a validated in vivo CT-based monitoring procedure for HIFU in fatty targets, and it comprises the following scientifically linked questions stated by the source work: (1) how blood perfusion, in vivo hydration, and tissue heterogeneity alter the HU-temperature trajectories in fatty tissue and the location or existence of the nadir; (2) which CT scanning protocol and radiation dose preserve detectability of the nadir and the quantitative fidelity of the HU-based indices while remaining clinically acceptable, the source explicitly calling for the lowest possible radiation dose to be found first in ex vivo studies and later in in vivo studies; and (3) whether the hysteresis-based indices (such as loop indentation-depth measures) reliably predict biologically meaningful damage metrics — cell viability, coagulation, histological necrosis — in vivo, with fat-containing clinical targets such as breast tumours constituting the motivating application for which the source states further work is needed to establish clinical monitoring applicability. The source additionally reports as a finding that the hysteresis phenomenon is tissue-type dependent (fat tracing a counter-clockwise loop versus the clockwise loops observed in liver and albumin phantoms), which suggests that distinct tissue types may require distinct calibrations or monitoring strategies — an optional extension of the objective rather than a mandatory sub-question. An admissible answer is a positive validation (a demonstrated in vivo protocol in perfused fatty tissue — animal models admissible, consistent with the source's staged ex vivo-then-in vivo program — with quantified trajectory fidelity, nadir detectability, dose report, and index-damage correlation) or a controlled refutation (in vivo evidence that the nadir or hysteresis features vanish, that the indices fail to predict damage, or that no clinically acceptable dose or scanning protocol preserves nadir detectability or index fidelity, under properly controlled conditions); these two forms are exhaustive.

References

  1. Weiss N, Sosna J, Goldberg SN, Azhari H (2014). Non-invasive temperature monitoring and hyperthermic injury onset detection using X-ray CT during HIFU thermal treatment in ex vivo fatty tissue. International Journal of Hyperthermia 30(2):119-125. DOI: 10.3109/02656736.2014.883466. https://doi.org/10.3109/02656736.2014.883466
  2. Weiss N, Goldberg SN, Sosna J, Azhari H (2014). Temperature-density hysteresis in X-ray CT during HIFU thermal ablation: Heating and cooling phantom study. International Journal of Hyperthermia 30(1):27-35. DOI: 10.3109/02656736.2013.860241. https://doi.org/10.3109/02656736.2013.860241
  3. Pandeya GD, Greuter MJ, de Jong KP, Schmidt B, Flohr T, Oudkerk M (2011). Feasibility of noninvasive temperature assessment during radiofrequency liver ablation on computed tomography. Journal of Computer Assisted Tomography 35(3):356-360. DOI: 10.1097/RCT.0b013e318217121d. https://doi.org/10.1097/RCT.0b013e318217121d
  4. Fani F, Schena E, Saccomandi P, Silvestri S (2014). CT-based thermometry: an overview. International Journal of Hyperthermia 30(4):219-227. DOI: 10.3109/02656736.2014.922221. https://doi.org/10.3109/02656736.2014.922221
  5. Pohlan J, Kress W, Hermann K-G, Mews J, Kroes M, Hamm B, Diekhoff T (2020). Computed tomography thermography for ablation zone prediction in microwave ablation and cryoablation: Advantages and challenges in an ex vivo porcine liver model. Journal of Computer Assisted Tomography 44(5):744-749. DOI: 10.1097/RCT.0000000000001081. https://doi.org/10.1097/RCT.0000000000001081
  6. Heinrich A, Schenkl S, Buckreus D, Güttler FV, Teichgräber UK-M (2022). CT-based thermometry with virtual monoenergetic images by dual-energy of fat, muscle and bone using FBP, iterative and deep learning-based reconstruction. European Radiology 32(1):424-431. DOI: 10.1007/s00330-021-08206-z. https://doi.org/10.1007/s00330-021-08206-z
  7. Kostyrko B, Rubarth K, Althoff C, Poch FGM, Neizert CA, Zibell M, Gebauer B, Lehmann KS, Niehues SM, Mews J, Diekhoff T, Pohlan J (2023). Computed tomography-based thermography (CTT) in microwave ablation: prediction of the heat ablation zone in the porcine liver. Insights into Imaging 14(1):189. DOI: 10.1186/s13244-023-01537-z. https://doi.org/10.1186/s13244-023-01537-z
  8. Koetzier LR, Hendriks P, Heemskerk JWT, van der Werf NR, Selles M, van der Molen AJ, Smits MLJ, Goorden MC, Burgmans MC (2025). Spectral computed tomography thermometry for thermal ablation: applicability and needle artifact reduction. Physica Medica 137:105093. DOI: 10.1016/j.ejmp.2025.105093. https://doi.org/10.1016/j.ejmp.2025.105093
  9. De Tommasi F, Massaroni C, Grasso RF, Carassiti M, Schena E (2021). Temperature monitoring in hyperthermia treatments of bone tumors: State-of-the-art and future challenges. Sensors 21(16):5470. DOI: 10.3390/s21165470. https://doi.org/10.3390/s21165470