A Monte Carlo detector response model for the IRIS PET preclinical scanner

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Biophysics

Background:

• Preclinical Positron Emission Tomography (PET) studies necessitate high spatial resolution for small animal imaging.
• Iterative image reconstruction algorithms are favored over analytical methods for their ability to model the entire imaging process comprehensively.

Purpose of the Study:

• To develop and validate an accurate factorized system matrix for the INVISCAN IRIS preclinical PET scanner.
• To improve image reconstruction in preclinical PET by incorporating detailed detector response modeling.

Main Methods:

• A novel factorized system matrix was created, comprising geometric and detector response components.
• The detector response matrix was generated using optimized Monte Carlo simulations to minimize computation time.
• The new system model was evaluated against the existing IRIS model using NEMA Standards NU 4-2008 criteria.

Main Results:

• The proposed system model demonstrated enhanced image quality compared to the current IRIS model.
• Improvements were observed in image uniformity and recovery coefficients, key metrics for PET performance.
• The inclusion of detector response in the system matrix significantly improved reconstruction outcomes.

Conclusions:

• The developed factorized system matrix accurately models the INVISCAN IRIS preclinical PET scanner.
• Incorporating detector response into the system matrix is crucial for achieving superior image reconstruction quality in preclinical PET.
• This approach offers a pathway to more precise and reliable small animal imaging studies.