4D maximum a posteriori reconstruction in dynamic SPECT using a compartmental model-based prior.

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Biomedical Engineering

Background:

• Dynamic SPECT imaging requires accurate reconstruction of time-varying activity distributions.
• Conventional methods like OSEM can suffer from artifacts and noise, impacting kinetic parameter estimation.
• Compartmental modeling is often used but can be limited by a priori assumptions.

Purpose of the Study:

• To introduce and evaluate a novel 4D ordered-subsets maximum a posteriori (OSMAP) algorithm for dynamic SPECT.
• To incorporate a temporal prior based on compartmental models directly into the image reconstruction process.
• To improve the accuracy and reduce uncertainty of kinetic parameter estimates in dynamic SPECT.

Main Methods:

• Developed a 4D OSMAP algorithm incorporating a temporal prior based on compartmental models.
• Applied the algorithm to dynamic 99mTc-teboroxime SPECT scans in patients and simulated phantom studies.
• Compared results with conventional ordered-subsets expectation-maximization (OSEM) followed by compartmental modeling.

Main Results:

• The dynamic OSMAP algorithm produced images with improved myocardial uniformity and definition.
• Time-activity curves showed reduced noise variations, and wash-in parameter estimates were more accurate with lower uncertainty.
• Bias in k21 estimates due to inconsistent projections was effectively removed for sampling schedules as slow as 60s.

Conclusions:

• The proposed dynamic OSMAP algorithm offers a flexible framework for dynamic tomographic imaging.
• It enhances image quality and kinetic parameter estimation in dynamic SPECT.
• This method has the potential to enable faster scanning protocols and improve noise reduction.