M.Sc Thesis

M.Sc StudentZiv-Ari Morris
SubjectReal-Time Control of Radio-Frequency Ablation by
DepartmentDepartment of Biomedical Engineering


Primary and secondary malignant hepatic tumors are some of the most common tumors, and represent a significant source of morbidity and mortality worldwide. Unfortunately, chemotherapy and radiation therapy are not very effective treatment methods.

Results from multiple investigations indicate that several minimally invasive treatment techniques are very effective for treating primary and secondary malignant hepatic tumors and have the potential of replacing surgical resection. Currently RF is the preferred technique for liver tumor ablation.

Different imaging techniques are used for the different stages of RF thermal ablation therapy - treatment planning, probe localization, and follow up of the treatment. Except for very few places that run the RF ablation procedure inside a MRI or CT machine, real-time control of the procedure is usually not done.

Several ultrasonic methods have been proposed to estimate temperature changes in tissue. These methods are related to different temperature-induced changes in the tissue, including changes in the attenuation, backscattered power, combination of speed of sound and thermal expansion, backscattered waveform change, or backscattered spectrum. Echo-strain estimation during ablation seems to be one of the promising techniques for temperature changes evaluation.

In this study, an algorithm was developed which is based on measurement of the spectral shift in the power spectrum (incoherent estimation technique) of the echo-signals. These signals, sampled from a commercial ultrasound system (VIVID 3, GE Healthcare), were used in this research for evaluating the echo-strains. Experiments were conducted in phantoms (tissue mimicking gel), in-vitro (turkey breast) and in-vivo (rabbit thigh). The phantom experiments showed positive frequency shifts as the temperature was elevated, with dependency on the heating rate. A linear relation (R2>0.96) was found in these experiments between the RF generator setting (imposed voltage) and the width of the treated area. In-vitro experiments showed similar results in homogeneous tissue regions. Further study has to be performed in order to enable evaluation of this algorithm in-vivo, since this method is very sensitive to movement of the examined area.