Investigation of CardioFreeze Motion Correction for Cardiac PET/CT

Cardiac imaging with positron emission tomography and computed tomography (PET/CT) is a principal technique in the diagnosis and prognostication of coronary artery disease. PET/CT image quality can be negatively impacted by patient, cardiac, and respiratory motion.

We evaluated images from 102 patients clinically referred for cardiac rest/stress Rb-82 PET/CT myocardial perfusion imaging. Patient images were processed with a prototype data-driven motion detection algorithm and stratified by motion severity. Images were subjected to cardiac (CS) or cardiac and respiratory (CD) motion correction via a novel reconstruction tool. Validated image quality markers including target-to-background ratio (TBR) were assessed between corrected and non-corrected images and analyzed via student's t-tests.

The motion detection algorithm identified 36 (35.3%) patients with significant motion. Image processing via CS yielded a statistically significant increase in mean image contrast (TBR 3.65±0.07, p<0.01) and intensified when respiratory motion correction was applied (3.78±0.07, p<0.005). Motion correction is more effective when applied to images containing greater motion, improving mean TBR by 3% (p<0.05) in non-motion scans but >10% (p<0.01) in significant motion cases. Three months post-imaging, 18 patients underwent clinically indicated coronary angiography: seven patients demonstrated significant motion, and CD motion correction improved their image quality to that of patients with no motion.

Over 1/3 of patients undergoing cardiac PET/CT have detectible motion leading to potential artifacts with clinical significance. Cardiac and respiratory motion correction improves image quality and may improve diagnostic accuracy, particularly when respiratory motion is severe. Further studies are critical to investigate the clinical impacts of improved image quality.