Improved image reconstruction of ⁸⁹Zr-immunoPET studies using a Bayesian penalized likelihood reconstruction...

Area of Science:

• Medical Imaging
• Nuclear Medicine
• Radiochemistry

Background:

• Positron Emission Tomography (PET) is crucial for molecular imaging.
• Zirconium-89 (89Zr) labeled immuno-PET tracers offer high sensitivity for detecting cancer.
• Optimizing image reconstruction is key to maximizing diagnostic accuracy and patient safety.

Purpose of the Study:

• To evaluate the Q.Clear Bayesian penalized likelihood algorithm for 89Zr-immunoPET image reconstruction.
• To assess the potential of Q.Clear to improve image quality.
• To determine if Q.Clear can reduce the required administered activity of 89Zr-immunoPET tracers.

Main Methods:

• Retrospective analysis of eight 89Zr-immunoPET whole-body PET/CT scans from clinical trials.
• Reconstruction of PET data using Q.Clear with varying beta values and standard OSEM.
• Simulation of reduced radiotracer activity by discarding 50% and 75% of counts.
• Quantitative and qualitative assessment of image quality and lesion detectability.

Main Results:

• Q.Clear reconstructions with a beta value of 3600 yielded the highest image quality scores.
• Images reconstructed with Q.Clear using 50% or 25% of original counts showed comparable or superior quality to standard OSEM with 100% counts.
• Signal-to-noise ratio and SUVmax measurements supported the improved image quality.

Conclusions:

• The Q.Clear algorithm significantly improves the image quality of 89Zr-immunoPET.
• Q.Clear facilitates the reduction of administered 89Zr-immunoPET activity.
• This advancement holds promise for enhancing future 89Zr-immunoPET studies and patient management.