Validation and Evaluation of a Vendor-Provided Head Motion Correction Algorithm on the uMI Panorama PET/CT System

Area of Science:

• Neuroimaging
• Medical Physics
• Radiochemistry

Background:

• Head motion (HM) in brain PET imaging introduces artifacts, reducing image resolution and impacting clinical applications like diagnosis and treatment planning.
• Accurate interpretation of PET scans is vital for neurological disease management.
• Existing methods may require manual parameter adjustments, complicating clinical workflows.

Purpose of the Study:

• To validate the efficacy of the NeuroFocus head motion correction (HMC) algorithm on the uMI Panorama PET/CT system.
• To assess the prevalence of head motion in short-duration clinical ¹⁸F-FDG brain PET scans.
• To evaluate the impact of HMC on image quality and diagnostic utility.

Main Methods:

• A data-driven, statistics-based HMC algorithm (NeuroFocus) was developed, automatically detecting HM without parameter adjustments.
• Validation involved 15 patients undergoing sequential 3-min ¹⁸F-FDG brain scans with and without induced motion.
• Evaluation included 302 clinical scans, categorizing HM based on SUV changes post-HMC and analyzing regional SUV_mean differences.

Main Results:

• NeuroFocus significantly reduced motion-induced differences in SUV_mean from -10.1% to -0.8% in the validation group.
• In the evaluation group, 12% of patients (38/302) exhibited significant HM requiring correction, with HMC improving SUV_mean by 10.9%.
• HMC enhanced image resolution and contrast, aiding disease identification and potentially reducing repeat scans.

Conclusions:

• The NeuroFocus algorithm effectively corrects head motion in short clinical ¹⁸F-FDG brain PET studies on the uMI Panorama PET/CT system.
• Approximately 12% of clinical brain PET scans benefit from HMC, establishing it as a valuable tool for improving diagnostic accuracy.
• Automated HMC optimizes clinical workflows by enhancing image quality and reducing the need for rescans.