Microwave thermal ablation using CT-scanner for predicting the variation of ablated region over time: advantages and limitations. (31st May 2019)

• Record Type: Journal Article
• Authors: Strigari, L; Minosse, S; D'Alessio, D; Farina, L; Cavagnaro, M; Cassano, B; Pinto, R; Vallati, G; Lopresto, V
• Is Part Of: Physics in medicine & biology. Volume 64:Number 11(2019:Jun.)
• Publication Date: 2019-05-31
• Subjects: microwave thermal ablation -- temperature prediction -- CT imaging -- treatment planning; Biophysics -- Periodicals; Medical physics -- Periodicals; 610.153
• DOI: 10.1088/1361-6560/ab1a67
• Languages: English
• ISSNs: 0031-9155

Abstract

Abstract: This study aims at investigating in real-time the structural and dynamical changes occurring in an ex vivo tissue during a microwave thermal ablation (MTA) procedure. The experimental set-up was based on ex vivo liver tissue inserted in a dedicated box, in which 3 fibre-optic (FO) temperature probes were introduced to measure the temperature increase over time. Computed tomography (CT) imaging technique was exploited to experimentally study in real-time the Hounsfield Units (HU) modification occurring during MTA. The collected image data were processed with a dedicated MATLAB tool, developed to analyse the FO positions and HU modifications from the CT images acquired over time before and during the ablation procedures. The radial position of a FO temperature probe (rFO) and the value of HU in the region of interest (ROI) containing the probe (HUo), along with the corresponding value of HU in the contralateral ROI with respect to the MTA antenna applicator (HUc), were determined and registered over time during and after the MTA procedure. Six experiments were conducted to confirm results. The correlation between temperature and the above listed predictors was investigated using univariate and multivariate analysis. At the multivariate analysis, the time, rFO and HUc resulted significant predictive factors of the logarithm of measured temperature. The correlation between predicted and measured temperatures was 0.934 ( p < 0.001). The developed tool allows identifying and registering the image-based parameters useful for predicting the temperature variation over time in each investigated voxel by taking into consideration the HU variation.