Motion detection in triple scan SPECT imaging

An algorithm to automatically detect and eliminate motion in projection data from triple scan SPECT imaging is proposed. The approach relies on a protocol where each head acquires projections over a full 360 degrees orbit rather than over the conventional 120 degrees partial orbit. Projection acquisition times are reduced to one third of that used for the partial orbit protocol to keep comparable total scan times. This protocol permits the acquisition of three full sets of SPECT data which can be adequately combined in order to reconstitute a motion-free set of projection data. Cross-correlation of the projection sets as a function of angle are used to detect motion. Plateaus of high correlation are estimated and matched in order to identify projections corresponding to the same position of a subject throughout a study. The motion correction algorithm was tested on experimental data obtained by manually moving a cardiac phantom at different times. In all cases, it was possible to reconstitute the evolution of the motion during the study. When no motion occurred during an interval equal to at least one third of the total acquisition time, high correlation plateaus were identified and complete sets of consistent projections were assembled from the longest matching plateaus. When more frequent motion occurred, an averaging of the three sets provided a means to improve the quality of the reconstructed slices compared to conventional imaging protocol.