Lesion detectability in motion compensated image reconstruction of respiratory gated PET/CT

Detectability of lung tumors in PET images is severely compromised by respiratory motion. In this work we investigate the performance of three different motion correction schemes for the PET data, with emphasis on lesion detectability. Motion information is derived by co-registration of the gated CT images. This is then used to compensate for the motion in the PET data by either deforming the images reconstructed without motion compensation, or by using the deformations as part of the system model of the reconstruction algorithm. We have previously investigated quantification [Manjeshwar, et al., 2006], Here, we extend this work and study the effect of three motion correction schemes on lesion detectability.