Normalization for 3D PET with a low-scatter planar source and measured geometric factors
1University of British Columbia/TRIUMF PET Centre, University Hospital, Vancouver, Canada.
Physics in Medicine and Biology
|May 8, 1998
Summary
A new 3D PET normalization method balances statistical accuracy and detector fidelity. This approach uses geometric factors (GFs) and detector efficiency to improve normalization, crucial for accurate positron emission tomography (PET) imaging.
Area of Science:
- Medical Imaging
- Nuclear Medicine
- Physics
Background:
- 3D Positron Emission Tomography (PET) normalization requires balancing statistical accuracy with detector/line-of-response (LOR) fidelity.
- Existing methods often compromise LOR accuracy for statistical quality or vice versa.
Purpose of the Study:
- To develop and implement a novel 3D PET normalization method for the ECAT 953B scanner.
- To evaluate the impact of algorithmic alterations on normalization accuracy and reconstructed image quality.
Main Methods:
- Developed a normalization method calculating detector normalization factors (NFs) as a product of geometric factors (GFs) and single detector efficiency (epsilon).
- Investigated the influence of GF matrix dimensionality and crystal positioning on normalization accuracy.
- Analyzed residual radial non-uniformities and their attribution to scatter and normalization interactions.
Main Results:
- Accurate GFs are critical; grouping similar LORs can lead to inaccurate NFs.
- Residual radial non-uniformities were observed across all tested NF variations.
- These non-uniformities are detectable even with low scan counts (13 million).
Conclusions:
- The developed 3D PET normalization method offers a viable approach for balancing statistical and detector fidelity.
- The method's generalizability to other tomographs is possible, with scanner-specific GF matrix dimensionality being key.
- Understanding scatter-normalization interactions is important for addressing residual non-uniformities in PET imaging.
