Optimization of ⁸⁹Zr PET Imaging for Improved Multisite Quantification and Lesion Detection Using an Anthropomorphic...

Area of Science:

• Nuclear Medicine
• Oncology Imaging
• Medical Physics

Background:

• Multicenter PET imaging presents challenges in data harmonization.
• Standardization is crucial for reliable oncology clinical trials.
• Zirconium-89 (89Zr) PET is valuable for targeted cancer imaging.

Purpose of the Study:

• To establish harmonization procedures for multicenter 89Zr PET imaging in oncology.
• To evaluate lesion detection and quantification across different PET scanners.
• To optimize image reconstruction for improved diagnostic performance.

Main Methods:

• Evaluation of seven PET scanners using a custom chest oncology phantom.
• Inclusion of nine spherical lesions (7-20 mm) with a 4:1 signal-to-background ratio.
• Assessment of lesion detection, recovery coefficients, and background signal variance with various reconstruction algorithms.

Main Results:

• Two of seven scanners did not meet performance standards.
• Optimal reconstruction algorithms were identified for the remaining five scanners.
• Average detection rate for 10-mm lesions was 95%; 7-mm lesions were visualized by only one scanner.
• Background variance ranged from 8.6% to 16%.

Conclusions:

• Multicenter harmonization procedures for 89Zr PET imaging in oncology were successfully established.
• Optimized protocols enhance the detectability of small lesions (≥10 mm).
• Accurate quantification of PET data is achievable across multiple centers with harmonization.