Abstract Archives of the RSNA, 2014
PHS147
Determining the Effect of Advanced Reconstruction Algorithms on Standardised Uptake Measurements of the Normal Adrenal Gland Using 18F-FDG PET/CT
Scientific Posters
Presented on December 1, 2014
Presented as part of PHS-MOB: Physics Monday Poster Discussions
Heok Cheow MBBCh, MSc, Presenter: Nothing to Disclose
Mark Gannon, Abstract Co-Author: Nothing to Disclose
Sarah Heard, Abstract Co-Author: Nothing to Disclose
There were significant difference between SUVmax measurements and visual appearance of the adrenal gland using the three reconstuction algorithms. Readers of PET/CT should be aware the normal range of adrenal uptake is dependant on reconstruction algorithm used.
Characteristation of adrenal incidentalomas found during 18F-FDG PET/CT imaging is important for clinical management. Many clinicians interpreting adrenal uptake use established SUV ranges of the normal adrenal gland or visually compared it with liver to aid their decision.
These SUV ranges were often determined from data processed using noncurrent iterative reconstruction algorithms. Contemporary commercial reconstruction utilise advanced algorithms which offer improved image quality (contrast, resolution) but may affect measured SUV. The purpose of the current study was to describe and compare normal adrenal 18F-FDG uptake between GE’s proprietary iterative reconstruction software VUE Point HD (V-HD) and their advanced VUE Point FX (V-FX) and VUE Point FX – Sharp IR (V-SIR) algorithms which utilise time of flight (TOF) and TOF with point spread function (PSF) corrections respectively.
A retrospective audit of 23 patients referred for lymphoma staging was performed with the expectation that adrenal uptake for this cohort would be representative of a normal patient population. All images were reconstructed using V-HD, V-FX and V-SIR algorithms. Quantitative analysis was performed by measuring SUVmax. Qualitative adrenal uptake was measured by visually comparing it to hepatic uptake (0 = no visualisation, 1 = activity less than liver, 2 = activity equal to liver, 3 = activity greater than liver).
There were statistically significant differences between reconstructed data for SUVmax and visual score results as determined by one-way ANOVA (F(2,99) = 32.16, p < 0.01 and F(2,99) = 9.69, p < 0.01 respectively). The range of SUVmax for all adrenal glands was 1.3 – 3.5, 1.5 – 4.5 and 2.5 – 7.9 for V-HD, V-IR and V-SIR respectively. Visual score correlated poorly with SUVmax (slope = 0.05, R = 0.04 slope = 0.24, R = 0.21 and slope = 0.83, R = 0.45) for V-HD, V-FX and V-SIR respectively.
Cheow, H,
Gannon, M,
Heard, S,
Determining the Effect of Advanced Reconstruction Algorithms on Standardised Uptake Measurements of the Normal Adrenal Gland Using 18F-FDG PET/CT. Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL.
http://archive.rsna.org/2014/14015457.html