Quantitative Assessment of Thermal Lesion Stiffness in the Liver: Initial ex vivo Results Hua Xie 1 , Shiwei Zhou 1 , Vijay Shamdasani 2 , Yan Shi 1 , Jean-Luc Robert 1 , John Fraser 2 , Shigao Chen 3 , and James F. Greenleaf 3 1 Philips Research North America, 345 Scarborough Road, Briarcliff Manor, NY 10510 2 Philips Healthcare, 22100 Bothell Everett Highway, Bothell, WA 98021 3 Department of Physiology and Biomedical Engineering, Mayo Clinic College of Medicine, Rochester, MN 55905 Abstract— The purpose of this study was to investigate the feasibility of a shear wave-based elastography prototype on Philips’ iU22 ultrasound system for measuring liver tissue stiffness before and after RFA (Radio Frequency Ablation) therapy. A special pulse sequence that generates and tracks shear waves was implemented on a curvilinear transducer C5-1, with RF data acquisition and off-line processing. The system was regulated as safe for use in human. Two RFA lesions were generated in ex vivo porcine liver by a commercial RF ablation electrode (Rita Medical Systems, CA). Lesions were created around depth 40 mm beneath the liver surface. The target temperature was set to 90°C and 95°C, and ablation lasted about 5 and 6 minutes for two lesions, respectively. Elasticity measurements were made pre-RFA at two sites where lesions were centered. Post RFA, the needle tines were retrieved to prevent interference with the shear wave propagation. To avoid residual bubbles, post-RFA elasticity measurement was performed at an image plane about 1 mm away in elevation from the ablation center plane containing the needle shaft. After each elasticity experiment, the liver was cut open approximately at the imaging plane to confirm the pathological changes caused by RFA. For statistical analysis, shear mechanical properties were estimated and averaged in a region with area about 10 mm x 4 mm at the center of the lesions. In this ex vivo study, the shear wave peak displacement at the push focus was about 10 μm pre- RFA. It was significantly reduced to 2 μm post-RFA, signaling the underlying elevated tissue stiffness. Quantitative reconstruction further confirmed the necrosis-induced stiffness increase. Pre-RFA, the shear modulus was 1.56 ± 0.19 kPa and 1.15 ± 0.25 kPa for lesion #1 and #2 respectively; post-RFA, the shear modulus was 29.48 ± 7.28 kPa and 26.80 ± 9.11 kPa. An increase of one order of magnitude in shear modulus was observed for both lesions. Similar stiffness contrast also existed between the ablated and neighboring non-treated region post- RFA for both lesions. These early results demonstrate the feasibility of the prototype for quantifying elasticity of thermal lesions in liver. Further technical development such as imaging capability and automated analysis tools is required to enable the visualization of the ablation zone and assessment of the therapy procedure. Keywords – RF ablation, ultrasound elastography, shear wave, shear modulus, RF electrode, RF needle I. INTRODUCTION Radio-frequency ablation (RFA) is gaining wide spread use in minimally invasive treatment of unresectable malignant hepatic tumors [1, 2]. The clinical problem associated with RFA procedures is the high local recurrence rate in treated patients: ranging from 1.7% to 41% for hepatocellular carcinoma [3]. One primary reason is that inaccurate identification of the ablation zone, especially errors in size measurement, leads to under-treatment. The success of ablation is currently measured by complete destruction of all malignant cells along with a safe margin of normal tissue. It is therefore important for physicians to be able to identify the ablation region and assess therapy success while the patient is still in surgery. A variety of imaging modalities have been attempted to visualize the ablation zone including conventional B-mode ultrasound (US), contrast-enhanced US (CEUS), contrast-enhanced CT, MRI and PET [4-5]. All of these modalities are of limited use because they carry certain disadvantages. Specifically, US does not provide adequate brightness contrast between ablated and surrounding tissue; the use of CEUS for liver imaging has not been approved by FDA, and the other modalities lack the capability of real-time evaluation and dynamic monitoring of the effects of therapy. During RFA procedures tissue stiffness is elevated due to thermal necrosis. The high stiffness contrast between treated and non-treated tissue has led to the development and use of ultrasound elastography for visualizing the coagulation zone during liver RFA. Both static compression based approaches and dynamic acoustic radiation force based approaches have been proposed [6-12]. We have previously demonstrated the feasibility of using real-time compression based strain imaging to assess liver RFA in in vivo animal studies [10]. In this study, we used a shear wave based ultrasound elastography prototype to quantitatively measure ex vivo liver stiffness pre and post- RFA. The prototype has been verified by measuring normal pig livers in vivo [13]. It has been further modified and regulated for safe human use. II. METHODS A. Shear Wave Elastography Prototype The shear wave elastography prototype was developed for a curvilinear transducer, C5-1, on a commercial ultrasound scanner iU22 (Philips Healthcare, Andover, MA). The scanner software was modified to implement a dedicated pushing and tracking pulse sequence. A typical pulse sequence consists of 10 identical segments of pushing and tracking events with a pulse repetition frequency (PRF) around 100 Hz. The entire pulse sequence lasts about 100 ms. The pushing pulse duration is a few hundred micro-seconds. It is focused at depth of 40 mm. Tracking pulse positions are evenly spaced to produce about 1 mm lateral beam spacing at the push focal depth. The