Aarhus University Seal / Aarhus Universitets segl

Lars Poulsen Tolbod

Effect of motion-induced PET-CT misalignment on cardiac function and myocardial blood flow measured using dynamic 15O-water PET

Research output: Contribution to conferenceConference abstract for conferenceResearchpeer-review

Aim: Motion-induced PET-CT misalignment artifacts are common in myocardial blood flow (MBF) measurements with 82Rb and 13N-ammonia. For 15O-water, MBF is based on the clearance rate rather than uptake of the tracer. The clearance rate is determined by the shape of the time-activity curve, not its amplitude, and is thus not affected by attenuation correction errors. Hence, misalignment is hypothesized not to affect 15O-water-based MBF to any large extent, but it may affect cardiac function measures derived from 15O-water scans. The aim of the present work was to assess the effect of PET-CT misalignment on MBF, transmural MBF (MBFt), perfusable tissue fraction (PTF), cardiac output (CO), stroke volume (SV) and left-ventricular ejection fraction (LVEF) based on dynamic 15O-water scans. Methods: 10 patients underwent 6 min PET scans after injection of 400 MBq 15O-water at rest and during adenosine-induced stress on a Discovery MI PET-CT scanner. Attenuation correction was based on a respiration- averaged CT scan. Images were reconstructed using TF-OSEM including PSF-recovery. PET-CT misalignment was induced by translation of the PET relative to the CT (1 cm caudal or cranial, 1 cm cranial + ventral, and 2.5 cm cranial + 1 cm ventral). MBF, MBFt, PTF, CO, SV and LVEF were calculated semi-automatically using aQuant software and values with and without misalignment were compared using correlation and Bland-Altman analysis. The original data, without misalignment, was analysed twice to assess intra-observer variability. Results: Intra-observer reproducibility was high, with R2≥0.98 for all parameters except LVEF during stress (R2 0.80). CO and SV were not significantly affected by misalignment, with correlation between aligned and misaligned values >0.95 (CO) and >0.93 (SV) and no significant bias for all motions. For MBF, correlation was always >0.95 for all coronary flow territories and motions, and no significant bias was seen. For MBFt however, correlations were always lower, reducing to 0.89 in LCX for the largest misalignment at stress. LVEF was only affected significantly for the largest misalignment. Correlation of misaligned PTF and original PTF values was lower, but no significant motion-induced bias was seen. Conclusion: With the exception of PTF and LVEF for >1.5 cm misaligment, misaligment- induced changes were generally similar to or slightly higher than intra-observer variability. PET-CT misalignment does not significantly affect MBF and cardiac function measures based on 15O-water scans for misalignment up to 1.5 cm.
Original languageEnglish
Publication year2017
Publication statusPublished - 2017
EventEuropean Association of Nuclear Medicine Annual Congress - Vienna, Austria
Duration: 21 Oct 201725 Oct 2017


ConferenceEuropean Association of Nuclear Medicine Annual Congress
Internet address

See relations at Aarhus University Citationformats

ID: 120302639