Analysis of RF Liver Ablation in Human Subjects Using Ultrasound Imaging, For The Prediction of Tumor Destruction

Tumor liver ablation suffers from different kinds of problems. One of them is the lack of real time monitoring of what happens to the tissue during the procedure and lack of knowledge when the tumor is being destroy.

Here, de-noising and analysis of the area being treated has been performed, using ultrasound clips from three different patients.
The de-noising was done using temporal wavelet de-noising and histogram adjustments. These stages provide good temporal and spatial results (CNR improved by 60%).

The analysis of the ablation has proved existence of two thermal bubble populations (p<0.05), with different amplitudes and concentrations.
A multi gaussian model has been employed on the data with good results (R-square0.96 , RMSE, verifying the existence of two populations. These results suggest the existence of a good predictor for the ablation success.