Dependence of Ultrasound Echo Decorrelation On Tissue Temperature During Radiofrequency Ablation of Ex Vivo Bovine Liver
S Subramanian1*, DT Schmidt1, TR Fosnight1, MB Rao2, TD Mast1, (1) Biomedical Engineering Program, University of Cincinnati, Cincinnati, OH, (2) Department of Environmental Health, University of Cincinnati, Cincinnati, OhioTU-E-144-2 Tuesday 2:00PM - 3:50PM Room: 144
Purpose: Echo decorrelation imaging is a ultrasound imaging method that has been shown effective for mapping of thermal ablation effects on liver and tumor tissue, both ex vivo and in vivo. Quantification of tissue temperature during thermal ablation treatments would benefit treatment guidance and control. Here, the effect of tissue temperature on the mapped echo decorrelation parameter is assessed for ex vivo radiofrequency ablation (RFA).
Methods: RFA treatments were performed on freshly acquired ex vivo bovine liver using a needle antenna with 500 kHz excitation for treatment voltages 30-40 V rms and durations 10-15 min. Throughout each treatment, beamformed ultrasound echo signals were recorded by a 7 MHz linear array and temperature was recorded by two thermocouples within the ultrasound image plane. Temperature maps were simulated by a finite element method, with tissue parameters determined using an unscented Kalman filter to best match measured ablation results. Echo decorrelation images were constructed from recorded echo data and correlated spatially with corresponding simulated temperature maps. Prediction of tissue temperatures over 60 °C, corresponding to local thermal ablation, was tested using receiver operating characteristic (ROC) curve analysis.
Results: Strong correspondence was observed between time-dependent echo decorrelation maps and corresponding temperature maps. Temperatures above 60 °C consistently corresponded to locally elevated echo decorrelation. Correlation coefficients between local temperature and echo decorrelation were weak but statistically significant. ROC curve analysis showed successful prediction of ablative tissue temperatures.
Conclusion: Observed correspondence between local tissue temperature and echo decorrelation indicates that elevated tissue temperatures consistently cause elevated echo decorrelation. Although weak overall correlation between local echo decorrelation and tissue temperature suggests that temperature is not accurately predicted by echo decorrelation, ROC curve analysis indicates successful prediction of tissue ablation, suggesting the utility of echo decorrelation imaging as a tool for real-time guidance and control of thermal ablation.
Funding Support, Disclosures, and Conflict of Interest: This research was supported by NIH grants R01 CA158439 and R21 EB008483.