Effects of motion compensation in CT perfusion

Computer Tomography perfusion (CTp) is arising as an important factor for prognostic evaluation of the effectiveness of antiangionetic therapies for different kinds of tumors. Respiratory movements can lead artifacts yielding possible inconsistent Time-Concentration Curves (TCCs), and unreliable perfusion values, accordingly. Despite several image-based motion compensation methods have been suggested in the literature, none of them measures the effect of motion on TCC signal, also correlating these measures to reliability of perfusion analyses. In this work, we study the effect of three rigid alignment approaches commonly used and with increasing quality, on both TCCs and perfusion maps in seven CTp studies of lung tumors, using a quantitative and a qualitative index. The results show that the expectedly best alignment yields the most likely perfusion maps. However, we first show that although the values of the quantitative index agree with the quality of alignments, it is not to be taken for granted that a better alignment yields a more likely perfusion pattern, according to the readers’ assessment