Abstract
1648
Objectives: Head motion is challenging for PET/SPECT studies that measure neuroreceptor BP using radioligands with slow kinetics. Scans require 2-3 hrs and restraints can produce discomfort that promotes head motion. Image-based realignment (SPM5) requires long frame durations (5-10 min) to obtain sufficient counts to delineate structural markers. Motion within frames causes small brain structures such as the basal ganglia (BG) to become blurred, decreasing measured activity. Shorter frames may limit within-frame motion, but counts may be insufficient for image-based realignment. This issue was addressed by comparing measured activity following image-dependent and image-independent motion correction in a PET study with short frame durations.
Methods: [F18]fallypride PET data were acquired in 1-min frames for 2.5 hrs while an infrared optical tracker (Polaris) recorded the sec-by-sec position of a target fixed to the participant’s head. For each frame, images were realigned to the target’s mean position and orientation while the corresponding standard deviations (SDs) quantified movement. TACs obtained from the BG were generated following motion correction using the optical-tracker or realignment with SPM5.
Results: Residual position uncertainty after motion correction with optical-tracker data was <0.5mm (rms), comparable to SPM5 for late frames when BG are prominent and well-defined, but not early frames when counts are low. When frames with large SDs were removed, mean measured activity increased and TACs were significantly smoother.
Conclusions: It is important to perform neuroligand studies with short frames to reduce blurring due to head motion. Optical tracking facilitates accurate realignment independent of frame duration and activity, quantifies within-frame head motion, and eliminates the need for uncomfortable head restraints.
- Society of Nuclear Medicine, Inc.