Related Experiment Videos
Effects of attenuation correction and reconstruction method on PET activation studies
Catalina T Mesina1, Ronald Boellaard, Odile A van den Heuvel
1PET Center, VU University Medical Center, Amsterdam, The Netherlands. c.mesina@vumc.nl
Neuroimage
|October 22, 2003
Summary
Ordered Subset Expectation Maximization (OSEM) iterative reconstruction reduces false positives in Positron Emission Tomography (PET) activation studies compared to filtered back-projection (FBP). OSEM offers similar results to FBP with reduced noise, making it a viable alternative.
Area of Science:
- Medical Imaging
- Neuroscience
- Radiology
Background:
- Statistical Parametric Mapping (SPM) analysis of Positron Emission Tomography (PET) activation studies relies on data reconstruction quality.
- Filtered Back-Projection (FBP) is standard, but iterative methods like Ordered Subset Expectation Maximization (OSEM) are gaining interest.
Purpose of the Study:
- To evaluate the impact of different PET data reconstruction techniques and attenuation correction (AC) on SPM activation detection.
- To compare OSEM and FBP methods regarding false positive rates and detection sensitivity.
Main Methods:
- Replicate and phantom studies assessed reconstruction effects on pixel variance and SPM outcomes.
- Volumetric analysis quantified false positives; average t-values were calculated.
- Clinical data from 11 patients were analyzed using SPM for all reconstruction types.
Main Results:
- Iterative OSEM reconstruction significantly reduced false positive voxels in both phantom and clinical data compared to FBP.
- OSEM provided comparable SPM results within activated areas to FBP.
- Attenuation correction (AC) improved noise for FBP but did not impact OSEM quality.
Conclusions:
- OSEM is a suitable alternative to FBP for PET activation studies, yielding SPM results with lower noise and fewer false positives.
- OSEM reconstruction maintains activation detection sensitivity while improving image quality.