Related Experiment Video
Updated: Jul 12, 2026
Computed Tomography-guided Time-domain Diffuse Fluorescence Tomography in Small Animals for Localization of Cancer Biomarkers
Published on: July 17, 2012
Advanced monitoring of TOF+PSF reconstruction performance in digital PET/CT: Clinical implications from the uMI 550
M El Kafhali1, M Tahmasbi2, A Rahmouni3
1Moulay Ismail University, Faculty of Sciences, Physical Sciences and Engineering, Innovative Research and Applied Physics (IRAP), Meknes, Morocco.
Introduction:
Advancements in digital PET/CT technology have enabled more precise quantification and image quality optimization. This study evaluated the reconstruction performance of the uMI 550 digital PET/CT system using standardized phantoms, focusing on reconstruction algorithms incorporating Time-of-Flight (TOF) and Point Spread Function (PSF) modeling. Unlike prior uMI 550 evaluations, this study integrates uncertainty-based analysis and longitudinal QC correlations to determine optimal reconstruction conditions.
Methods:
Performance assessments followed NEMA NU 2-2007 protocols with a NEMA IEC image quality phantom and QC phantoms. The analysis examined contrast recovery, image noise, and standardized uptake value (SUV) metrics (SUVmax (single hottest voxel), SUVmean (average uptake in a region), and SUVpeak (average uptake within a fixed 1 cm3 around the hottest area)), under varying reconstruction settings, with integration of uncertainty modeling and quality-control (QC) correlations. Count-rate performance was assessed using Noise Equivalent Count Rate (NECR) and true count rate.
Results:
Contrast recovery rose from 89.1 % at iteration 1-152.2 % at iteration 4, after which further iterations increased noise without added benefit. SUVmean remained stable (1.01 ± 0.03), SUVpeak converged rapidly with minimal sensitivity to reconstruction, while SUVmax overshot to 1.34 by iteration 6 due to edge artifacts. Count-rate analysis showed peak NECR at 17.2 kBq/mL and maximum true count rate of 479.4 kcps at 26.3 kBq/mL, with quantification errors consistently below 5 %. Correlation analysis (r > 0.85) identified SUVpeak, SUVmean, NECR, and contrast recovery as the most influential QC parameters.
Conclusion:
SUVpeak emerged as the most robust quantitative biomarker, providing practical guidance for protocol optimization, harmonized PET/CT workflows, and long-term system monitoring.
Implications For Practice:
Adopting evidence-based reconstruction strategies and consistent QC can improve diagnostic confidence and standardization in clinical PET/CT imaging.
