Intercostal high intensity focused ultrasound for liver ablation: The influence of beam shaping on sonication efficacy and near-field risks

M. De Greef, G. Schubert, J. W. Wijlemans, J. Koskela, L. W. Bartels, C. T.W. Moonen, M. Ries

Research output: Contribution to journalArticlepeer-review

16 Citations (Scopus)


Purpose: One of the major issues in high intensity focused ultrasound ablation of abdominal lesions is obstruction of the ultrasound beam by the thoracic cage. Beam shaping strategies have been shown by several authors to increase focal point intensity while limiting rib exposure. However, as rib obstruction leaves only part of the aperture available for energy transmission, conserving total emitted acoustic power, the intensity in the near-field tissues inherently increases after beam shaping. Despite of effective rib sparing, those tissues are therefore subjected to increased risk of thermal damage. In this study, for a number of clinically representative intercostal sonication geometries, modeling clinically available hardware, the effect of beam shaping on both the exposure of the ribs and near-field to acoustic energy was evaluated and the implications for the volumetric ablation rate were addressed. Methods: A relationship between rib temperature rise and acoustic energy density was established by means of in vivo MR thermometry and simulations of the incident acoustic energy for the corresponding anatomies. This relationship was used for interpretation of rib exposure in subsequent numerical simulations in which rib spacing, focal point placement, and the focal point trajectory were varied. The time required to heat a targeted region to 65 °C was determined without and with the application of beam shaping. The required sonication time was used to calculate the acoustic energy density at the fat.muscle interface and at the surface of the ribs. At the fat.muscle interface, exposure was compared to available literature data and rib exposure was interpreted based on the earlier obtained relation between measured temperature rise and simulated acoustic energy density. To estimate the volumetric ablation rate, the cool-down time between periods of energy exposure was estimated using a time-averaged power limit of 100 kJ/h. Results: At the level of the ribs, the temperature density proportionality constant was estimated to be °C/(J/mm2). Beam shaping by the geometric shadow method typically reduces the acoustic intensity a factor of 2, considering the 1 cm2 with the highest exposure. For a 4 mm diameter circular sonication trajectory, the near-field energy limit of 2.5 J/mm2 was exceeded for all considered geometries. The estimated rib temperature was in all but one (sonication 50 mm behind the ribs, with 15 mm rib spacing and a 4 mm diameter circular sonication trajectory) of the considered scenarios within acceptable limits. For those sonication scenarios where a single sonication is considered safe both in terms of near-field as well as rib heating, volumetric ablation rates in the order of 1 ml/h are estimated. Conclusions: Intercostal sonication is associated with an increased risk of near-field overheating. This risk is strongly dependent on the considered rib spacing, the placement of the focus behind the ribs, and the selected sonication trajectory. For the hardware under simulation, obstruction by the thoracic cage renders ablations of clinically relevant volumes within a practical time-frame unfeasible in a large part of the liver. Improvements maybe expected from transducer designs with a larger active surface and/or nonlinear sonication strategies.

Original languageEnglish
Pages (from-to)4685-4697
Number of pages13
JournalMedical Physics
Issue number8
Publication statusPublished - 1 Aug 2015
Externally publishedYes


  • high-intensity focused ultrasound
  • near-field heating
  • rib obstruction
  • simulation


Dive into the research topics of 'Intercostal high intensity focused ultrasound for liver ablation: The influence of beam shaping on sonication efficacy and near-field risks'. Together they form a unique fingerprint.

Cite this