Related Experiment Video
Updated: Jul 15, 2026
10:31
Non-invasive Imaging and Analysis of Cerebral Ischemia in Living Rats Using Positron Emission Tomography with 18F-FDG
Published on: December 28, 2014
Partial volume effect compensation for quantitative brain SPECT imaging
Yong Du1, Benjamin M W Tsui, Eric C Frey
1Department of Radiology, The Johns Hopkins University, Baltimore, MD 21287, USA. duyong@jhu.edu
IEEE Transactions on Medical Imaging
|August 12, 2005
Summary
This study introduces a new partial volume compensation method for brain SPECT imaging, significantly improving quantitative accuracy. The perturbation-based geometric transfer matrix (pGTM) method reduces underestimation of brain activity and ratios.
Area of Science:
- Medical Imaging
- Nuclear Medicine
- Quantitative Analysis
Background:
- Partial volume (PV) effects in SPECT brain imaging reduce quantitative accuracy.
- Existing PV compensation (PVC) methods need adaptation for iterative reconstruction in SPECT.
Purpose of the Study:
- To extend the geometric transfer matrix (GTM) method for PV compensation in brain SPECT.
- To adapt the GTM method for iterative reconstruction, developing a perturbation-based GTM (pGTM).
Main Methods:
- Extended the GTM method for brain SPECT using iterative reconstruction.
- Developed the perturbation-based GTM (pGTM) by treating region maps as perturbations.
- Evaluated pGTM using Monte Carlo simulations and experimental SPECT data.
Main Results:
- PVC using pGTM significantly improved quantitative accuracy in brain SPECT.
- Reduced underestimation of striatal activities from 30% to 1.2% in simulated data.
- Reduced underestimation of striatal activities from 36% to 7.8% in experimental data.
Conclusions:
- The pGTM method effectively compensates for partial volume effects in brain SPECT.
- This advancement enhances the reliability of quantitative SPECT imaging for clinical applications.
