Motion-corrected echo decorrelation imaging of in vivo focused and bulk ultrasound ablation in a rabbit liver cancer model

Abstract

Echo decorrelation imaging, a pulse-echo method that maps heat-induced changes in ultrasound echoes, was investigated for in vivo monitoring of thermal ablation in a liver cancer model. In open surgical procedures, rabbit liver with VX2 tumor was imaged by 64-element image-ablate arrays and sonicated at 5.00, 5.05, or 5.20 MHz by unfocused 32- or 64-element apertures for bulk ultrasound ablation and electronically focused 64-element apertures for focused ultrasound ablation. Echo decorrelation and integrated backscatter (IBS) images were formed from pulse-echo signals recorded during rest periods following each sonication pulse. Echo decorrelation images were corrected for motion- and noise-induced artifacts using measured echo decorrelation from corresponding sham trials. Sectioned ablated tissue was vitally stained with triphenyl tetrazolium chloride (TTC) and binary images were constructed based on local TTC uptake. Echo decorrelation was significantly greater in ablated regions than in non-ablated regions. Motion correction significantly reduced echo decorrelation in non-ablated regions. Prediction of cell death by echo decorrelation and IBS imaging was assessed using receiver operating characteristic (ROC) curves. Areas under the ROC curve (AUROC) were significantly greater than chance for corrected and uncorrected echo decorrelation as well as IBS. Corrected echo decorrelation predicted ablation significantly better than IBS for the focused ultrasound exposures. AUROC differences between corrected echo decorrelation and IBS were not statistically significant for the unfocused exposure group, for which the IBS AUROC was marginally higher, or for the combination of all exposures, for which the corrected echo decorrelation AUROC was marginally higher. These results confirm that echo decorrelation imaging can effectively predict local thermal ablation in vivo.