MIRD Pamphlet No. 32: A MIRD Recovery Coefficient Model for Resolution Characterization and Shape-Specific...

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Radiotherapy Physics

Background:

• Accurate quantification in emission tomography is vital for internal radiopharmaceutical therapy dosimetry.
• The partial-volume effect significantly impacts measurements in objects smaller than 10 times the imaging system's spatial resolution.
• Current methods often struggle with accurate correction for non-spherical objects.

Purpose of the Study:

• To develop a framework for characterizing spatial resolution in PET and SPECT.
• To propose novel MIRD recovery coefficient models (RECOVER-GM and RECOVER-EM) for shape-specific partial-volume correction (PVC).
• To validate these models and assess their effectiveness in reducing quantification errors.

Main Methods:

• Developed a framework for PET and SPECT spatial resolution characterization.
• Proposed two MIRD recovery coefficient models: geometric mean approximation (RECOVER-GM) and an empirical model (RECOVER-EM).
• Validated models using simulations and phantom experiments, including ellipsoidal phantoms for comparative PVC testing.

Main Results:

• The RECOVER models demonstrated significant error reduction in activity quantification compared to conventional sphere-based corrections.
• Error reduction factors ranged from approximately 1.3 to 5.7 in comparative PVC tests on ellipsoidal phantoms.
• The proposed methodology proved effective for region-based PVC, including corrections for nonspherical volumes.

Conclusions:

• The developed RECOVER models and PVC methodology offer a robust framework for accurate activity quantification in emission tomography.
• These tools enhance region-based PVC, particularly for nonspherical targets like tumors.
• This work contributes to the MIRDsoft.org project, aiming to improve dosimetry tools for disease characterization and treatment planning.