Study of different Time Attenuation Curve processing in liver CT perfusion

Perfusion CT is a feasible technique for obtaining quantitative informations about tumor-related vascularization of liver. Time Attenuation Curve (TAC) computation is the first step to obtain perfusion parameters from mathematical modeling. We apply “slope method” analysis to TAC to assess arterial blood flow (BFa) in liver tumor (HCC-Hepatocellular Carcinoma). A new algorithm which removes outlier points of TAC was proposed to correct motion artifacts that affect Hounsfield Unit measurement. Two different procedures were then applied to fit the original data points of tumor TAC. Moreover, in this study we compared three methods of TAC processing (a spline smoothing interpolation, a gamma fitting, no fitting with analysis on raw data) in the evaluation of BFa in liver tumors. A group of eight HCC patients was analyzed to evaluate the BFa tumor variability related to these different approaches. The results of this work demonstrated that TAC data processing influences the assessment of BFa in HCC patients and that smoothing spline interpolation is a good processing method to reduce irregularities of tumor liver TAC, preserving its basic shape.