Research ArticleClinical Investigations

Variability and Uncertainty of 18F-FDG PET Imaging Protocols for Assessing Inflammation in Atherosclerosis: Suggestions for Improvement

Pauline Huet, Samuel Burg, Dominique Le Guludec, Fabien Hyafil and Irène Buvat
Journal of Nuclear Medicine April 2015, 56 (4) 552-559; DOI: https://doi.org/10.2967/jnumed.114.142596

Article Figures & Data

Figures

  • FIGURE 1.
    FIGURE 1.

    (A and B) Numeric model of 2 simulated lesions. (C) Activity map of modified XCAT phantom including atherosclerotic lesion.

  • FIGURE 2.
    FIGURE 2.

    Patient (A) and simulated (B) images (OSEM algorithm, 1-mm voxel size, PSF modeling). Arrows point to lesions.

  • FIGURE 3.
    FIGURE 3.

    Sample images obtained for same lesion acquired and reconstructed using different protocols (50 ENIs). Images that look most pleasant visually due to low noise are not the ones in which SUV estimate is most accurate (true SUV = 14.1).

  • FIGURE 4.
    FIGURE 4.

    Large lesion. Mean error in SUV estimate (expressed as percentage of true SUV, scale on left of graph) as function of SD of that mean percentage error over 10 replicates. Results are shown for SUVmax (A), SUVmean (B), and SUVmeanofmax (C) and for 6 different protocols. dX = acquisition duration of X minutes; vY = voxel size = Y mm; f, with a 4-mm in FWHM postfiltering included; psf, with a 4-mm in FWHM PSF modeling included. Colored arrows indicate direction of increasing iterations (1 point every 10 ENIs), showing that increasing iteration number reduces average percentage error. Corresponding estimated SUVs are given by scale on right of each graph.

  • FIGURE 5.
    FIGURE 5.

    Small lesion. Mean error in SUV estimate (expressed as percentage of true SUV, scale on left of graph) as function of SD of that mean percentage error over 10 replicates. Results are shown for SUVmax (A), SUVmean (B), and SUVmeanofmax (C) and for 6 different protocols. dX = acquisition duration of X minutes; vY = voxel size = Y mm; f, with a 4-mm in FWHM postfiltering included; psf, with a 4-mm in FWHM PSF modeling included. Colored arrows indicate direction of increasing iterations (1 point every 10 ENIs), showing that increasing iteration number reduces average percentage error. Corresponding estimated SUVs are given by scale on right of each graph.

Tables

  • TABLE 1

    List of Abbreviations Used in Manuscript

    AbbreviationMeaning
    BABlood activity
    dX-vY protocolProtocol involving an X-min duration acquisition with reconstructed voxel size of Y mm
    dX-vY-f protocoldX-vY protocol, followed by 3-dimensional gaussian postfiltering of 4 mm in FWHM in each direction
    dX-vY-psf protocoldX-vY protocol involving a gaussian kernel model of 4 mm in FWHM within the reconstruction
    ENIEffective number of iterations
    IcountsIndex characterizing the number of counts in the raw data to be reconstructed
    LBRLesion–to–blood activity ratio
    OSEMOrdered-subset expectation maximization
    OWAObserved wall activity
    PSFPoint spread function
    PVEPartial-volume effect
    RWAReal wall activity
    SUVStandardized uptake value
    SUVmaxMaximum SUV in the volume of interest
    SUVmeanMean SUV in the volume of interest
    SUVmeanofmaxMean of the SUVmax measured in all transaxial slices encompassing the lesion
    TBRTarget–to–blood activity ratio
    TBRmaxTBR calculated using the maximum uptake in the target
    TBRmeanTBR calculated using the mean uptake in the target
    TBRmeanofmaxTBR calculated by averaging the maximum uptake measured in all transaxial slices encompassing the target
    VOIVolume of interest
  • TABLE 2

    Parameters Systematically Recorded in All Articles as Part of Bibliographic Study

    Image acquisitionReconstructionQuantitative analysis
    Acquisition systemVoxel size*Measured metric*
    Injected activity*Spatial resolution in reconstructed images*VOI size
    Postinjection timeReconstruction algorithmVOI delineation method
    Acquisition duration*Effective number of iterations*PVE correction
    Investigated arterial bedPostfiltering*Metric values
    Attenuation correction
    Scatter correction
    Resolution recovery*

    • ↵* We investigated influence of parameters noted with asterisk using simulated data (see “Assessment of Impact of Various Acquisition/Reconstruction/Analysis Procedures” section).

  • TABLE 3

    Activity Values Assigned to Different Compartments of XCAT Phantom

    LabelActivity (kBq/mL)
    Blood3.3
    Healthy adventitia, intima, media1.6
    Inflammatory adventitiaLBR × 3.3
    Atheroma1.4
    Esophagus6.3
    Salivary gland4.0
    Thyroid3.7
    Bone marrow3.3
    Parathyroid2.5
    Body0.6
    Muscle1.6
    Rib6.3
    Skin3.2
    Brain19.7
    Spinal cord7.16
    Laryngopharynx0.60

Additional Files

Back to top

In this issue

Print
Download PDF
Article Alerts
Email Article
Citation Tools
Bookmark this article

Jump to section

Keywords