Original Article
Enhanced definition PET for cardiac imaging

https://doi.org/10.1007/s12350-010-9193-7Get rights and content

Abstract

Background

We aimed to determine in phantom and cardiac clinical studies the impact of a new high-resolution PET image reconstruction.

Methods

A phantom with cardiac insert filled with 18F, 14 18F-FDG viability studies and 15 82Rb perfusion studies were acquired on a Siemens Biograph-64 (4-ring). The data were reconstructed with 2D- and 3D-attenuation weighted ordered subsets expectation maximization (AWOSEM), and high-definition reconstruction (HD·PET). We calculated wall/cavity contrast, contrast-to-noise ratio (CNR), wall thickness, motion/thickening and ejection fraction.

Results

In the phantom study, we found an increase in defect size (up to 26%), contrast (up to 48%) and CNR (1.9) with HD·PET as compared to standard techniques. The contrast increased on HD·PET images compared to 2D- and 3D-AWOSEM for viability (14.0% ± 4.8%) and perfusion studies (7.3% ± 4.3%) (P < .05). Average CNR increased with HD·PET by 79.4% ± 17.1% and 68.8% ± 3.0% in viability and perfusion studies respectively (all P < .05). Average wall thickness with HD·PET decreased in the phantom study by 1.3 ± 0.3 mm and the viability studies by 1.9 ± 0.7 mm but not in the perfusion studies. The functional measurements were not significantly different for any techniques.

Conclusions

We demonstrated both in phantom and patient cardiac studies that HD·PET improves image contrast, defect definition, and CNR.

Introduction

Reconstruction techniques have an impact on image quality and quantitative analysis in cardiac PET. Recently, new high-resolution reconstruction techniques which take advantage of resolution recovery principles have been presented.1, 2, 3, 4 In particular, high definition reconstruction for PET (HD·PET) has been introduced by Siemens Healthcare Molecular Imaging (Knoxville, TN).5 Unlike the classic reconstruction techniques, this new iterative reconstruction algorithm compensates for the distortions introduced in the final images by the circular geometry of the scanner. The result is an image with improved spatial resolution and a better control of the noise.5

In this work, we hypothesized that HD·PET provides cardiac images with a higher technical quality and we aimed to verify this in a cardiac phantom and in patient images. We tested the impact of HD·PET on image contrast, noise and myocardium wall thickness. We compared HD·PET reconstruction to regular 2D and 3D reconstruction techniques used at our institution. Phantom experiments were performed with Fluorine-18 and patient images were obtained with Fluorine-18 fluorodeoxyglucose (18F-FDG)6 and perfusion studies were obtained using Rubidium-82.7 We show significant improvements in technical cardiac image quality with the use of the high-resolution reconstruction.

Section snippets

PET Acquisition and Reconstruction

All images were acquired on a Siemens Biograph-64 TruePoint PET/CT with the TrueV option. This 3D system consists of a 64-slice CT and a PET scanner with 4 rings of lutetium oxyorthosilicate (LSO) detectors with a detector element dimensions of 4 × 4 × 20 mm3.8 The image plane spacing is 2 mm. The PET axial and transaxial FOV are 216 and 605 mm, respectively. The coincidence time window and the energy window are respectively 4.5 ns and 425-650 keV. The data was acquired in list mode format. A

Reconstruction Time

For all the reconstructions, we used Windows XP PC with two 2.33 GHz Intel Xeon processors. After the list mode sorting phase, the reconstruction time of the static images (matrix size 168 × 168 × 109) was 1 minute 57 seconds for 2D-AWOSEM (3 iterations 8 subsets, 24 ML equivalent iterations), 1 minute 58 seconds for 3D-AWOSEM (3 iterations 14 subsets, 42 ML equivalent iterations) and 2 minutes 35 seconds for HD·PET (4 iterations and 14 subsets, 56 ML equivalent iterations). The gated

Discussion

HD·PET modeling of the 3D detector spatial response during the reconstruction step is designed to better control the noise in the reconstructed images in general.5 In this work, we have shown that HD·PET improves significantly the technical image quality of cardiac studies. The noise, contrast and wall thickness were improved in phantom, viability studies and perfusion studies (static and gated).

The change in contrast and wall thickness can in most part be explained by the improved spatial

Conclusions

Our work on both phantom and patient cardiac studies showed that HD·PET significantly improved defect definition, image resolution, contrast and contrast-to-noise and provided equivalent functional quantitative information compared to standard reconstruction techniques. This improved technical quality could lead to improved performance of cardiac PET/CT. Further evaluation of the clinical impact of HD·PET on cardiac imaging would appear to be warranted.

Acknowledgments

The authors thank Jimmy Fermin and Brandi N. Huber from Cedars-Sinai Medical Center for their help with the PET acquisition, Heidi Gransar for her help with the statistical analysis and Michael Casey from Siemens Healthcare for his help with the setting up of the High Definition.

References (21)

  • ManjeshwarR et al.

    Fully 3D PET iterative reconstruction using distance-driven projectors and native scanner geometry

    IEEE Nucl Sci Symp Conf Rec

    (2006)
  • ChangG et al.

    Comparison between two super-resolution implementations in PET imaging

    Med Phys

    (2009)
  • KennedyJ et al.

    Super-resolution in PET imaging

    IEEE Trans Med Imaging

    (2006)
  • De BernardiE et al.

    Resolution recovery in PET during AWOSEM reconstruction: A performance evaluation study

    IEEE Trans Nucl Sci

    (2007)
  • PaninVY et al.

    Fully 3-D PET reconstruction with system matrix derived from point source measurements

    IEEE Trans Med Imaging

    (2006)
  • KudoT

    Metabolic imaging using PET

    Eur J Nucl Med Mol Imaging

    (2007)
  • Di CarliM et al.

    Clinical myocardial perfusion PET/CT

    J Nucl Med

    (2007)
  • JonssonC et al.

    A comparison of the imaging properties of a 3-and 4-ring biograph PET scanner using a novel extended NEMA phantom

    IEEE Nucl Sci Symp Conf Rec

    (2007)
  • TownsendD et al.

    Performance and clinical workflow of a new combined PET/CT scanner

    J Nucl Med

    (2007)
  • WatsonC et al.

    Prompt gamma correction for improved quantification in 82Rb PET

    J Nucl Med

    (2008)
There are more references available in the full text version of this article.

Cited by (43)

  • Technical Aspects of Cardiac PET Imaging and Recent Advances

    2016, Cardiology Clinics
    Citation Excerpt :

    This resolution recovery method has been equivalently termed point spread function or high-definition reconstruction, and is currently offered by all commercial PET vendors. Effective tomographic image resolution as low as 2 mm and significant improvements in image contrast have been achieved using resolution recovery16; better definition of subtle myocardial perfusion defects have been reported in clinical settings with the use of this approach17 (Fig. 11). The combination of TOF reconstruction and point spread function correction has also been shown to improve image quality.18,19

  • Recent Advances and Future Progress in PET Instrumentation

    2016, Seminars in Nuclear Medicine
    Citation Excerpt :

    The penalized likelihood iterative reconstruction was designed to reach convergence of the iterative reconstruction without amplifying noise in the reconstruction while maintaining a uniform resolution recovery across the FOV (Fig. 6). Effective tomographic image resolution as low as 2 mm (in the air) and significant improvements in image contrast have been achieved using resolution recovery,20 and better definition of subtle myocardial perfusion defects have been reported in clinical settings with the use of this approach,21 as shown in Figure 7. The combination of TOF reconstruction and PSF correction has also been shown to improve image quality.22,23

  • Advances in SPECT and PET Hardware

    2015, Progress in Cardiovascular Diseases
    Citation Excerpt :

    The combination of the TOF reconstruction and correction of the resolution recovery showed the best improvement of image quality.34,35 Clinically, better definition of subtle myocardial perfusion defects has been reported with the use of the resolution recovery.36 The advanced image reconstructions utilized for 3D PET are significantly different from those used previously in 2D PET.

  • New cardiac cameras: Single-photon emission CT and PET

    2014, Seminars in Nuclear Medicine
    Citation Excerpt :

    In particular, the combination of the TOF and correction of the point spread function seems to result in the best improvement of image quality.46,47 Cardiac PET imaging with the correction of the point spread function has been reported, demonstrating better definition of subtle defects.48 An example of improved image resolution for 18F FDG imaging with the use of resolution recovery is shown in Figure 20.

View all citing articles on Scopus
View full text