Evaluation of motion-correction methods for dual-gated cardiac positron emission tomography/computed tomography
Riku Klén1, Tommi Noponen, Juha Koikkalainen
1aTurku PET Centre, Turku University Hospital bDepartment of Mathematics and Statistics cDepartment of Cell Biology and Anatomy, Institute of Biomedicine, University of Turku dDepartment of Nuclear Medicine eDepartment of Medical Physics, Turku University Hospital, Turku fKnowledge Intensive Services, VTT Technical Research Centre of Finland, Tampere, Finland gHammersmith Imanet Ltd hInstitute of Nuclear Medicine, University College London, London, UK.
Background:
Dual gating is a method of dividing the data of a cardiac PET scan into smaller bins according to the respiratory motion and the ECG of the patient. It reduces the undesirable motion artefacts in images, but produces several images for interpretation and decreases the quality of single images. By using motion-correction techniques, the motion artefacts in the dual-gated images can be corrected and the images can be combined into a single motion-free image with good statistics.
Aim:
The aim of the present study is to develop and evaluate motion-correction methods for cardiac PET studies. We have developed and compared two different methods: computed tomography (CT)/PET-based and CT-only methods.
Methods:
The methods were implemented and tested with a cardiac phantom and three patient datasets. In both methods, anatomical information of CT images is used to create models for the cardiac motion.
Results:
In the patient study, the CT-only method reduced motion (measured as the centre of mass of the myocardium) on average 43%, increased the contrast-to-noise ratio on average 6.0% and reduced the target size on average 10%. Slightly better figures (51, 6.9 and 28%) were obtained with the CT/PET-based method. Even better results were obtained in the phantom study for both the CT-only method (57, 68 and 43%) and the CT/PET-based method (61, 74 and 52%).
Conclusion:
We conclude that using anatomical information of CT for motion correction of cardiac PET images, both respiratory and pulsatile motions can be corrected with good accuracy.
