Research ArticleInstrumentation

Evaluation of Sinus/Edge-Corrected Zero-Echo-Time–Based Attenuation Correction in Brain PET/MRI

Jaewon Yang, Florian Wiesinger, Sandeep Kaushik, Dattesh Shanbhag, Thomas A. Hope, Peder E.Z. Larson and Youngho Seo
Journal of Nuclear Medicine November 2017, 58 (11) 1873-1879; DOI: https://doi.org/10.2967/jnumed.116.188268

Article Figures & Data

Figures

  • FIGURE 1.
    FIGURE 1.

    Workflow of generating pseudo-CT from ZTE images: ZTACSEC and ZTACUC. In the binary mask, false-positive bone pixels were corrected to stay below 42 HU (soft tissue).

  • FIGURE 2.
    FIGURE 2.

    Localization of 20 VOIs for VOI-based quantification: axial brain (A), coronal brain (B), axial cerebellum (C), sagittal brain with sinus and cerebellum (D); left-right (1–2) thalamus, (3–4) corpus callosum, (5–6) temporal lobe, (7–8) frontal lobe, (9–10) occipital lobe, (11–12) cerebral cortex, (13–20) cerebellum at anterior (13–14), posterior (15–16), inferior (17–18), and center (19–20) positions.

  • FIGURE 3.
    FIGURE 3.

    Bland–Altman plot of 240 VOIs of 12 patients for PET uptake differences (%) with MRAC (ZTACUC, ZTACSEC, ATAC) from PET with CTAC (PET/CTAC). PET/ZTACSEC (red stars) was close to zero mean difference mostly within ±5% boundaries.

  • FIGURE 4.
    FIGURE 4.

    Box plots of PET uptake differences (%) for 12 patients across 20 VOIs (A) and for 20 VOIs across 12 patients (B) between PET/CTAC and PET/MRAC (left, blue: ZTACUC; middle, red: ZTACSEC; right, black: ATAC). Tops and bottoms of each box are 25th and 75th percentiles of samples, respectively, with median and outlier (+ sign, >1.5 × interquartile range).

  • FIGURE 5.
    FIGURE 5.

    Histograms of voxel-based PET uptake differences (PET/MRAC − PET/CTAC) across all patients using voxels within SUV range of 0.5–15.0 (g/mL). (A) ZTACUC. (B) ZTACSEC. (C) ATAC. Counts were normalized by maximum count of PET/ZTACUC difference from PET/CTAC. (Left to right) Yellow lines indicate 25th and 75th percentiles.

  • FIGURE 6.
    FIGURE 6.

    Patients 5 (A and B) and 8 (C and D) in coronal and sagittal views for CT and pseudo-CTs (first row), attenuation-correction (AC) maps (second row), and corresponding PET images (third row) according to AC methods. Second and third rows illustrate difference images on second through fourth columns for pseudo-CT–derived AC maps from CT-derived AC maps (unit, cm−1) and PET/MRAC from PET/CTAC (unit, SUV), respectively. Arrows indicate sinus, mastoid, and cerebellum.

Tables

  • TABLE 1

    PET Uptake Differences Between Reference PET/CTAC and PET/MRAC (ZTACUC, ZTACSEC, and ATAC) Across All 240 VOIs

    Difference
    PET/MRACkBq/mL%P
    PET/ZTACUC0.79 ± 0.525.6 ± 3.5<0.01
    PET/ZTACSEC0.03 ± 0.330.2 ± 2.40.23
    PET/ATAC−0.10 ± 0.71−0.9 ± 5.00.03

    • Differences are mean ± SD. A paired t test was performed for pair of PET/CTAC and PET/MRAC.

Back to top

In this issue

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

Jump to section

Keywords