Original Article |
| |
Efficacy of motioncorrection in absolute quantification of colonic PET-CT for drug response therapyMichael A Masoomi, Andy Robinson, Yassine Bouchareb, Seyedali Hejazi, Nicholas M Spyrou. Abstract | | | | Aims: The study aimed at: a) characterizing and correcting bowelmotion induced artefacts whilst imaging the region in pre and post drug therapy 18FFDG scans; b)developing a motion model of the gut using a fully 3D nonrigid registration technique for applying to NACT digitised images. Methods: A motion correction technique for PET-CT scans, particularly those of the abdomen and colon was developed. Attenuation and activity image volumes were generated at different points in the respiratory cycle using the Nonuniform Rational Bspline Cardiac Torso (NCAT) anthropomorphic phantom. The movement of the abdomen was characterised as part of the image registration process and assessment of the motion correction technique was performed quantitatively with Region of Interest (ROI), image fidelity, and image correlation techniques; semiquantitatively with line profile analysis; and qualitatively by overlaying non-motion-corrected and motion-corrected image frames. Results: Motion correction was successful for frames that were substantially different to the reference (large motion); these frames had considerable differences between the ROI activities in the non-motion-corrected and reference frames. Large motion correction resulted in an improvement in image fidelity factor (from 0.848 to 0.976). Conclusion: In principle, PET-CT motion correction of the colon can be performed using image registration between different frames in the respiratory cycle. Clinically, frames at different points in the respiratory cycle can be obtained by respiratory gating during PET image acquisition. Future work can concentrate on developing this technique so that it can be applied to clinical data.
|
|
|
|