Echo decorrelation imaging of ex vivo HIFU and bulk ultrasound ablation using image-treat arrays

Abstract

In this study, the ability of ultrasound echo decorrelation imaging to map and predict heat-induced cell death was tested using bulk ultrasound thermal ablation, high intensity focused ultrasound (HIFU) thermal ablation, and pulse-echo imaging of ex vivo liver tissue by a custom image-treat array. Tissue was sonicated at 5.0 MHz using either pulses of unfocused ultrasound (N=12) (7.5 s, 50.9-101.8 W/cm2 in situ spatial-peak, temporal-peak intensity) for bulk ablation or focused ultrasound (N=21) (1 s, 284-769 W/cm2 in situ spatial-peak, temporal-peak intensity and focus depth of 10 mm) for HIFU ablation. Echo decorrelation and integrated backscatter (IBS) maps were formed from radiofrequency pulse-echo images captured at 118 frames per second during 5.0 s rest periods, beginning 1.1 s after each sonication pulse. Tissue samples were frozen at −80˚C, sectioned, vitally stained, imaged, and semi-automatically segmented for receiver operating characteristic (ROC) analysis. ROC curves were constructed to assess prediction performance for echo decorrelation and IBS. Logarithmically scaled mean echo decorrelation in non-ablated and ablated tissue regions before and after electronic noise and motion correction were compared. Ablation prediction by echo decorrelation and IBS was significant for both focused and bulk ultrasound ablation. The log10-scaled mean echo decorrelation was significantly greater in regions of ablation for both HIFU and bulk ultrasound ablation. Echo decorrelation due to electronic noise and motion was significantly reduced by correction. These results suggest that ultrasound echo decorrelation imaging is a promising approach for real-time prediction of heat-induced cell death for guidance and monitoring of clinical thermal ablation, including radiofrequency ablation and HIFU.