Comparison of ultrasound strain images with multi-modality imaging techniques in liver RF ablation assessment: Initial ex vivo and clinical results

Ultrasound (US) elastography in liver imaging applications has recently gained interest in the clinical community. Various techniques explore the use of elasticity information for clinical diagnosis (for fibrosis-cirrhosis staging) and assessment of radiofrequency (RF) or High Intensity Focused Ultrasound (HIFU) ablation. In this paper, we conduct compression-based strain imaging for ex vivo and clinical liver RF ablation (RFA) assessment, using the RFA needle to induce the strain (method initially proposed by Varghese et al. in ex vivo porcine). The results are validated against several imaging modalities. Six ex vivo bovine livers were embedded in gelatin mixtures. A commercial RFA needle (Rita) was inserted and deployed into each liver sample. Radiofrequency (RF) ultrasound data were collected using Philips US systems during RFA needle pulling. Strain images were generated offline using 2D speckle tracking. Post- RFA strain images were compared to T1-weighted MRI (Philips). The livers were then dissected along the US image plane for caliper measurements. The elasticity images of the lesions exhibit a 40-fold contrast-to-noise Ratio (CNR) enhancement as compared to B-mode. The lesion sizes, as measured on the elasticity images, compared well with the MRI measurements (<10% error). Clinical RF data were acquired on two RFA patients with an iU-22 system (Philips). Both patients underwent pre- and post-operative contrast computed tomograpy (CT) imaging, and one also had intra-operative contrastenhanced ultrasound. Following RFA, no lesion could be clearly identified in the US B-mode images. Comparable lesion dimensions were measured from the intra-operative elasticity images and post-operative contrast CT. Elasticity imaging could provide similar information as contrast CT or contrast ultrasound imaging without the need for contrast or additional ionizing radiation. It also potentially allows for intra-operative monitoring of the RFA procedure, which in turn could lead to improved efficacy of RFA treatments.