Quantitative accuracy of MAP reconstruction for dynamic PET imaging in small animals

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Image Reconstruction

Background:

• Iterative reconstruction algorithms are increasingly used in positron emission tomography (PET) imaging.
• Quantitative accuracy of reconstructed PET images needs validation across varying contrast and counting statistics.

Purpose of the Study:

• Evaluate the quantitative accuracy of the 3D maximum a posteriori (3D-MAP) algorithm for dynamic PET imaging.
• Compare 3D-MAP against 2D filtered-backprojection (2D-FBP) and 2D-ordered subsets expectation maximization (2D-OSEM).

Main Methods:

• Utilized Siemens microPET scanners for dynamic PET imaging studies.
• Employed specially designed phantoms to assess spatial resolution, image quality, and quantitative accuracy.
• Evaluated blood activity concentration accuracy in a mouse model by comparing image-derived values with arterial blood sampling.

Main Results:

• 3D-MAP achieved superior spatial resolution (~1.0 mm) compared to 2D-FBP (1.8 mm) and 2D-OSEM (1.6 mm).
• 3D-MAP demonstrated better quantitative accuracy and higher recovery coefficients, especially for smaller structures.
• A minor positive bias was noted in 3D-MAP for very low count densities, and accurate blood activity concentration prediction was limited by myocardial spill-over in dynamic mouse studies.

Conclusions:

• 3D-MAP provides more accurate activity concentration estimates than 2D-FBP and 2D-OSEM in PET imaging.
• Improved spatial resolution with 3D-MAP leads to reduced partial volume effects and enhanced quantitative accuracy.
• The study validates 3D-MAP as a robust reconstruction algorithm for dynamic PET imaging applications.