The heart position shifts considerably due to motion associated with the re
spiratory cycle, and this motion can degrade the image quality of cardiac-g
ated positron emission tomography (PET) studies. One method to combat this
motion-induced blur is a respiratory-gated acquisition followed by recombin
ation of registered image volumes using a rigid-body motion assumption; how
ever, nonrigid deformation of the heart from respiratory motion may reduce
the effectiveness of this procedure. We have investigated a 12-parameter gl
obal affine motion model for registration of different respiratory gates in
an end-diastolic cardiac PET sequence. To obtain robust estimates of motio
n, a four-dimensional registration model was devised that encouraged smooth
ly varying motion between adjacent respiratory time frames. Registration pa
rameters were iteratively calculated using a cost function that combined a
least squares voxel difference measure with a penalty obtained from a predi
ction prior. The prior was calculated from adjacent time frames assuming co
nstant velocity and an affine model. After registration, the principal exte
nsion ratios were calculated to measure the degree of nonrigid motion. In d
ata from ten subjects, extension ratios of over 5% were common, indicating
that an affine model may provide better registrations and in turn, better m
otion-corrected composite volumes than could a technique restricted to the
six-parameter rigid body assumption.