Image Reconstruction for a Partially Collimated Whole Body PET Scanner

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Image Reconstruction

Background:

• Partially collimated PET systems offer potential count rate advantages over conventional 2-D and 3-D systems.
• Previous research has not definitively established image-based improvements or optimized reconstruction methods for partially collimated data.

Purpose of the Study:

• To develop and evaluate an image reconstruction method specifically designed for partially collimated PET data.
• To address the lack of customized reconstruction techniques for these systems.

Main Methods:

• Developed a tailored image reconstruction method using simulated and measured sensitivity patterns.
• Incorporated a measured normalization correction term to address unique true event sensitivities.
• Proposed a modified scatter correction based on simulated data.

Main Results:

• Measured image quality data validated the effectiveness of the normalization correction for true events.
• The modified scatter correction was found to be unnecessary based on image quality assessments.

Conclusions:

• The proposed reconstruction method, particularly the normalization correction, is effective for partially collimated PET data.
• Further optimization of scatter correction may not be required for achieving image quality improvements.