PURPOSE: As quantitative F-FDG PET numbers and pooling of results from different PET/CT scanners become more influential in the management of patients, it becomes imperative that we fully interrogate differences between scanners to fully understand the degree of scanner bias on the statistical power of studies. PATIENTS AND METHODS: Participants with body mass index (BMI) greater than 25, scheduled on a time-of-flight (TOF)-capable PET/CT scanner, had a consecutive scan on a non-TOF-capable PET/CT scanner and vice versa. SUVmean in various tissues and SUVmax of malignant lesions were measured from both scans, matched to each subject. Data were analyzed using a mixed-effects model, and statistical significance was determined using equivalence testing, with P < 0.05 being significant. RESULTS: Equivalence was established in all baseline organs, except the cerebellum, matched per patient between scanner types. Mixed-effects method analysis of lesions, repeated between scan types and matched per patient, demonstrated good concordance between scanner types. CONCLUSIONS: Patients could be scanned on either a TOF or non-TOF-capable PET/CT scanner without clinical compromise to quantitative SUV measurements.
PURPOSE: As quantitative F-FDG PET numbers and pooling of results from different PET/CT scanners become more influential in the management of patients, it becomes imperative that we fully interrogate differences between scanners to fully understand the degree of scanner bias on the statistical power of studies. PATIENTS AND METHODS: Participants with body mass index (BMI) greater than 25, scheduled on a time-of-flight (TOF)-capable PET/CT scanner, had a consecutive scan on a non-TOF-capable PET/CT scanner and vice versa. SUVmean in various tissues and SUVmax of malignant lesions were measured from both scans, matched to each subject. Data were analyzed using a mixed-effects model, and statistical significance was determined using equivalence testing, with P < 0.05 being significant. RESULTS: Equivalence was established in all baseline organs, except the cerebellum, matched per patient between scanner types. Mixed-effects method analysis of lesions, repeated between scan types and matched per patient, demonstrated good concordance between scanner types. CONCLUSIONS:Patients could be scanned on either a TOF or non-TOF-capable PET/CT scanner without clinical compromise to quantitative SUV measurements.
Authors: Urs J Muehlematter; Hannes W Nagel; Anton Becker; Julian Mueller; Kerstin N Vokinger; Felipe de Galiza Barbosa; Edwin E G T Ter Voert; Patrick Veit-Haibach; Irene A Burger Journal: EJNMMI Res Date: 2018-05-31 Impact factor: 3.138