RT Journal Article SR Electronic T1 Which breathing protocol is best for computer-based fusion of independent PET and CT data sets? JF Journal of Nuclear Medicine JO J Nucl Med FD Society of Nuclear Medicine SP 406P OP 406P VO 48 IS supplement 2 A1 Aleksandar Grgic A1 Dirk Hellwig A1 Ursula Nestle A1 Andrea Schaefer-Schuler A1 Stephanie Kremp A1 Carl Martin Kirsch YR 2007 UL http://jnm.snmjournals.org/content/48/supplement_2/406P.4.abstract AB 1706 Objectives: To determine the optimum breathing protocol for the fusion of separate 18F-FDG-PET and CT studies of the thorax. Methods: Six patients (5m, mean 64,5 years, 53-71 years) were examined on separate PET- and CT- scanners on the same day positioned identically (verified by positioning aid, laser skin marks and photo documentation). CTs were performed using different breathing protocols (expiration, inspiration, mid breathhold) in every patient. The data sets were coregistrated with a non-rigid algorithm (normalized mutual information) on a commercially available workstation (Hermes Medical Solutions). Three fusion groups were formed: Group 1: CT-Inspiration with PET, Group 2: CT-Expiration with PET, and Group 3: CT-Mid breathhold with PET. The quality of the fusion was assessed in each group based on the following anatomical landmarks: thyroid, lung apices, aortic arch, heart, spine, carina, liver dome/diaphragm and tumor. To rate the differences between the groups a scale from 1 (no alignement) to 5 (exact alignement) was used. To compare these results a Wilcoxon signed rank test was used. Results: Comparing the groups group 3 yielded a better fusion of the data (mean 4,3) as group 1 (3,7) or group 2 (3,1). Group 3 showed in five of eight anatomical landmarks a significantly better alignment, as these were: liver dome (p=0,03), heart (p=0,04), tumor (p=0,03), lung apices (p=0,03), and spine (p=0,02), respectively. Furthermore, group 3 provided better results in comparison to the group 2 with respect to carina (p=0,04) and the tumor (p=0,02), respectively. Conclusions: In this preliminary investigation the fusion beween the separate PET- and CT-studies obtained at mid breathhold showed the least deviation. Thus, this protocol is applied in our daily routine.