Lars Poulsen Tolbod

Work flow with manual re-registration of cardiac 18F-FDG PET and attenuation correction CT to compensate for respiratory motion: Effects and inter-operator variability.

Research output: Contribution to conferencePosterResearchpeer-review

  • Christian Juhl
  • Nana Louise Christensen, Nuklearmedicinsk Afdeling og PET-center, Denmark
  • Lone W. Møller, Nuklearmedicinsk Afdeling og PET-center, Denmark
  • Kirsten Bouchelouche, Institut for Klinisk Medicin, Nuklearmedicinsk Afdeling og PET-center, Denmark
  • Lars Poulsen Tolbod
Aim: Cardiac PET exams with CT for attenuation correction are increasing in number. However, due to different natures of the PET (average over many respiratory cycles) and the CT exam (snap-shot during one respiratory cycle) and the large difference in density of the cardiac and lung tissue, attenuation artefacts are often observed on the lateral wall of the left ventricle. Many strategies has been proposed to reduce these artefacts, including averaging multiple CTs, respiratory gating and manual re-registration1-3. However, for scanner platform-independent implementation, manual re-registration is currently the most feasible. In this study, we examine both the effect on misalignment artefacts and the inter-operator variability of manual re-registration performed by the technologist in a step before the final image reconstruction and verified by a second technologist before image analysis. Method: 40 18F-FDG viability scans were examined. After the patient scan, a fast image reconstruction of low quality was made and used for manual re-registration of PET and AC CT, before starting the high quality reconstruction used for analysis. The left ventricle was reoriented and segmented in QPS (Cedar-Sinai) and quantified using AHA-17-scores. Comparisons were made between 1) no re-registration, 2) re-registration by two independent technologists. Differences of 10% on the segment level were considered significant 2-3. Results and conclusion: When no manual re-registration was performed, significant artefacts were observed in 15 % (6/40) of the scans, and, as expected, mainly in lateral segments. It was estimated that 80% (5/6) of these artefacts could be reduced using manual re-registration procedure. This is in agreement with previous reports1-3. The inter-operator variability was generally low, except in a few cases where the manual re-registration resulted in new artefacts. These were, however, easily identified and eliminated by second technologist in the verification-step resulting in an overall insignificant inter-operator variability. In conclusion, manual re-registration reduces artefacts due to misalignment in 18F-FDG viability studies and can be performed with insignificant inter-operator variability. We suggest that manual re-registration is performed routinely in daily clinical practise combined with control by a second person. References:1)Kennedy J.A. et al., J Nucl Med 2009; 50:1471-1478, 2)Lautamäki R. et al., Eur J Nucl Mol Imaging 2008;35:305-310, 3)Martinez-Möller A. et al., J Nucl Med 2007;48:188-193
Original languageEnglish
Publication year2013
Publication statusPublished - 2013
EventEANM '13 - Lyon, France
Duration: 19 Oct 201323 Oct 2013


ConferenceEANM '13

See relations at Aarhus University Citationformats

ID: 76800459