The feasibility of controlling high-intensity focused ultrasound (HIFU) thermal ablation in real time using echo decorrelation imaging feedback was investigated in ex vivo bovine liver. Sonication cycles (5.0 MHz, 0.7… Click to show full abstract
The feasibility of controlling high-intensity focused ultrasound (HIFU) thermal ablation in real time using echo decorrelation imaging feedback was investigated in ex vivo bovine liver. Sonication cycles (5.0 MHz, 0.7 s per HIFU pulse, 20-24% duty, 879-1426 W/cm2 spatial-peak, and temporal peak intensity) performed by a linear image-ablate array were repeated until the minimum cumulative echo decorrelation within the focal region of interest exceeded a predefined threshold. Based on preliminary experiments (N = 13), a threshold of −2.7 for the log10-scaled echo decorrelation per millisecond was defined, corresponding to 90% specificity of local ablation prediction. Controlled HIFU thermal ablation experiments (N = 10) were compared with uncontrolled experiments employing 2, 5, or 9 sonication cycles. Controlled trials showed significantly smaller average lesion area (4.78 mm2), lesion width (1.29 mm), and treatment time (5.8 s) than 5-cycle (7.02 mm2, 1.89 mm, 14.5 s) or 9-cycle (9.31 mm2, 2.4 mm, 26.1 s) uncontrolled trials. Prediction of local ablation using echo decorrelation was assessed using receiver operator characteristic (ROC) curve analysis, in which controlled trials showed significantly greater prediction capability (area under the ROC curve AUC = 0.956) compared to 2-cycle uncontrolled trials (AUC = 0.722). These results suggest that ablation control using echo decorrelation may improve the precision, reliability, and duration of ultrasound-guided HIFU treatments.The feasibility of controlling high-intensity focused ultrasound (HIFU) thermal ablation in real time using echo decorrelation imaging feedback was investigated in ex vivo bovine liver. Sonication cycles (5.0 MHz, 0.7 s per HIFU pulse, 20-24% duty, 879-1426 W/cm2 spatial-peak, and temporal peak intensity) performed by a linear image-ablate array were repeated until the minimum cumulative echo decorrelation within the focal region of interest exceeded a predefined threshold. Based on preliminary experiments (N = 13), a threshold of −2.7 for the log10-scaled echo decorrelation per millisecond was defined, corresponding to 90% specificity of local ablation prediction. Controlled HIFU thermal ablation experiments (N = 10) were compared with uncontrolled experiments employing 2, 5, or 9 sonication cycles. Controlled trials showed significantly smaller average lesion area (4.78 mm2), lesion width (1.29 mm), and treatment time (5.8 s) than 5-cycle (7.02 mm2, 1.89 mm, 14.5 s) or 9-cycle (9.31 mm2, 2.4 mm, 26.1 s)...
               
Click one of the above tabs to view related content.