Abstract
1938
Objectives Acceptance testing of a novel solid state cardiac camera identified the potential for artefacts to be introduced when there is a difference in the position of the heart in the FOV between stress and rest imaging >1cm. The objective of this study was to develop a method of aligning stress and rest images to an accuracy of <=1cm.
Methods Software was developed to display the previously acquired stress images overlaying the persistence rest images during patient set-up. Adjustment of contours around the heart and blending between the two images allows the operator to visualise any misalignment and patient position to be adjusted accordingly; facilitating accurate manual alignment. A 30s scout scan is acquired and software incorporating the Sheffield Image Registration Toolkit (ShIRT) is employed to calculate the difference in the position of the heart at stress and rest, along with the bed and gantry positions required for alignment. The patient is moved into the recommended position and the rest images are acquired. Registration is again employed to calculate any residual misalignment between the stress and rest images at the post acquisition stage. The scan is repeated if the final misalignment is >1cm.
Results Review of 53 consecutive scans (43 patients) has shown that using this technique scans were aligned to an accuracy of 0.4cm +/- 0.2cm (1SD), with a maximum difference 0.9cm.
Conclusions Software has been developed that not only facilitates quick and easy alignment of the stress and rest images on the GE Discovery NM 530c cardiac gamma camera but also facilitates image QC following the scan to quantify any remaining misalignment. This allows scans to be repeated where necessary. Using this technique stress and rest images can be aligned to an accuracy of <=1cm