Evaluation of a breath-motion-correction technique in reducing measurement error in hepatic CT perfusion imaging

Objective: To evaluate the effect of a breath-motion-correction (BMC) technique in reducing measurement error of the time-density curve (TDC) in hepatic CT perfusion imaging. Methods: Twenty-five patients with suspected liver diseases underwent hepatic CT perfusion scans. The right branch of portal vein was selected as the anatomy of interest and performed BMC to realign image slices for the TDC according to the rule of minimizing the temporal changes of overall structures. Ten ROIs was selected on the right branch of portal vein to generate 10 TDCs each with and without BMC. The values of peak enhancement and the time-to-peak enhancement for each TDC were measured. The coefficients of variation (CV) of peak enhancement and the time-to-peak enhancement were calculated for each patient with and without BMC. Wilcoxon signed ranks test was used to evaluate the difference between the CV of the two parameters obtained with and without BMC. Independent-samples t test was used to evaluate the difference between the values of peak enhancement obtained with and without BMC. Results: The median (quartiles) of CV of peak enhancement with BMC [2.84% (2.10%, 4.57%)] was significantly lower than that without BMC [5.19% (3.90%, 7.27%)] (Z=-3.108,P<0.01). The median (quartiles) of CV of time-to-peak enhancement with BMC [2.64% (0.76%, 4.41%)] was significantly lower than that without BMC [5.23% (3.81%, 7.43%)] (Z=-3.924, P<0.01). In 8 cases, TDC demonstrated statistically significant higher peak enhancement with BMC (P<0.05). Conclusion: By applying the BMC technique we can effectively reduce measurement error for parameters of the TDC in hepatic CT perfusion imaging. (authors)