NetForum uses cookies to ensure that we give you the best experience on our website. If you continue to use the site, we'll assume that you are happy to receive these cookies on the NetForum website. Read about our cookies.
NetForum Community
Learn. Share. Optimize.
Log in | Sign up now | Submit content | Contact
Go to similar content

Lesion detectability and effective spatial resolution with LOR reconstruction

White Paper
Philips NM Marketing Philips Healthcare

GEMINI GXL system introduces a new data reconstruction method called Line of Response (LOR) reconstruction. LOR reconstruction is the next generation fully 3D reconstruction method that builds on a Philips tradition to provide the most advanced 3D reconstruction algorithms. LOR reconstruction is the most accurate data reconstruction technique available today and it provides high effective spatial resolution and improved lesion detectability. In a traditional (non-LOR) method, raw data is rearranged prior to reconstruction and combined to form a uniform, interpolated sinogram. This necessary interpolation step includes a "filtering" effect that degrades lesion detectability and spatial resolution (Figure 1). LOR reconstruction eliminates the interpolation step to preserve the maximum spatial resolution of the system. The performance of the GEMINI GXL scanner with and without LOR reconstruction was evaluated by measuring the lesion detectability in an IEC phantom and calculating the effective spatial resolution of the system.
Fig. 1: The
Fig. 1: The "filtering" effect of traditional, non-LOR reconstruction
Raw sinogram (left), before interpolation, and the same data after interpolation to uniform sampling. The loss in resolution from the interpolation step (right) should translate to an equivalent loss of resolution in the reconstructed image.

Fig. 2: LOR vs. non-LOR reconstruction methods
Fig. 2: LOR vs. non-LOR reconstruction methods

Figure 2 shows the IEC phantom images for a visual comparison between LOR and non-LOR iterative reconstruction methods. 

 

IEC phantom images with LOR reconstruction demonstrate:

  • Higher contrast in the smaller spheres
  • More accurate (round) shape of the spheres
  • Improved background uniformity

 

 

Fig. 3:  Lesion detectability metric LOR reconstruction has better lesion detectability as compared to non-LOR iterative reconstruction and to FBP. This result is interpreted as an improvement in the effective spatial resolution. For the same lesion detectability value, a smaller object can be visualized with LOR reconstruction (compared to non-LOR and FBP).
Fig. 3: Lesion detectability metric
LOR reconstruction has better lesion detectability as compared to non-LOR iterative reconstruction and to FBP. This result is interpreted as an improvement in the effective spatial resolution. For the same lesion detectability value, a smaller object can be visualized with LOR reconstruction (compared to non-LOR and FBP).

Figure 3 shows a measured lesion detectability metric, which is the ratio of the hotspot amplitude to the amplitude of background noise of similar frequency.

 

The lesion detectability metric reflects two independent variables responsible for the visual identification of lesions:

  • Contrast recovery
  • Background noise

 

Images with the same noise but higher contrast recovery will have better lesion detectability. The same is true for the images with the same contrast recovery and lower background noise. 

 

PET system spatial resolution is typically meausured using the NEMA Nu-2 2001 protocol. This is a measurement of a point source response. In this test, a small (< 1 mm) point source and filtered back-projection reconstruction is used to generate an image. 

 

Typical GEMINI GXL spatial resolution (1 cm from center, radial) measured using the NEMA protocol is 5.2 mm.  However, this measurement does not take into account the benefits of LOR reconstruction because NEMA test protocol specifies that a FBP-based reconstruction technique be used.  Therefore, we introduce the parameter effective spatial resolution, which includes the improvement in spatial resolution from LOR reconstruction. Effective spatial resolution is defined as the product of NEMA point-source spatial resolution and relative lesion detectability improvement. Using this definition, the typical effective spatial resolution of the GEMINI GXL scanner is 4.5 mm (1 cm from center, radial).

Summary

The LOR reconstruction technique improves the effective spatial resolution of the scanner. This performance improvement should result in better detection of small lesions and improved diagnostic confidence.

 

References

  • D.J. Kadrmas, "LOR-OSEM: statistical PET reconstruction from raw line-of-response histograms." Phys. Med. Biol. 49 (2004) 4731-4744.
  • M. Rafecas, et al. "Use of a Monte-Carlo based Probability Matrix for 3D Iterative reconstruction of MADPET - II Data." Conference Record, IEEE MIC Conference, Portland, OR 2003.
  • R.J. Walledge, et al. "Quad-HIDAC PET: Comparison of Four Image Reconstruction Techniques for High Resolution imaging." Conference Record, IEEE MIC Conference, Rome Italy, October 16-22, 2004.
  • D. Gagnon, J. Kolthammer, Physics and Reconstruction of GEMINI GXL. Philips Report, May 2005.


This content has been made possible by NetForum Community.
Share this on: Share your link in twitter Share your link in facebook Share your link on LinkedIn Print Rate this article: Log in to vote

 
Rating:
Votes:
0
Views:
5241
Added:
Apr 26, 2010

Rate this:
Log in to vote
 

White Paper
Gemini GXL / LXL systems
3D, FBP reconstruction, lesion detectability, lesions, Oncology
 

Clinical News
Best Practices
Case Studies
Publications and Abstracts
White Papers
Web seminars and Presentations
ExamCards
Protocols
Application Tips and FAQ
Training
Try an Application
Business News
Case Studies
White Papers
Web Seminars and Presentations
Utilization Services
Contributing Professionals
Contributing Institutions
Become a Contributor