Cardiac motion compensation improves reproducibility of 18F-sodium fluoride PET uptake quantification in the aortic valve

Objectives Combined positron emission tomography (PET) and computed tomography (CT) using 18F-sodium fluoride (18F-NaF) radioactive tracer has been recently shown to predict disease progression and clinical outcome in patients with aortic stenosis. Previous studies have used data reconstructed from a single diastolic gate (1/4th of the cardiac cycle) to avoid blurring due to cardiac motion at the expense of image noise. We examined a novel software motion correction method which allows the use of data from the full cardiac cycle in order to improve the reproducibility of quantitative tracer uptake in the aortic valve.

Methods Methods: Fifteen patients (73±7 y/o) with aortic stenosis underwent two combined 18F-NaF (125 MBq) PET and CT angiography scans within two months. ECG-gated PET images were reconstructed with time-of-flight and resolution recovery, and fused with CCTA. A variational mass-preserving image registration algorithm was applied to align PET images from 4 cardiac gates representing full cardiac cycle. Maximal Standard Uptake Value (SUVmax) measurements in the aortic valve were computed for the diastolic gate images alone with 25% of the counts, cardiac motion corrected images, and ungated data (data from full cardiac cycle without motion correction) for both sets of scans.

Results The SUVmax difference (mean ± standard deviation) between the 2 repeated studies was 0.31±0.37 for ungated data, 0.30±0.21 for the diastolic gate, and 0.24±0.19 for motion-corrected data, all with P<0.001. The bias and 95% limits of agreement for the SUVmax differences by Bland-Altman analysis of the difference between the 2 repeated studies were 0.19, (-0.68 to 1.01) for ungated study, 0.07 (-0.65 to 0.8) for diastolic gate, and 0.035 (-0.58 to 0.65) for the motion corrected data.

Conclusions The use of nonlinear motion correction to compensate for cardiac motion in 18F-NaF aortic valve PET data improves the reproducibility of SUVmax measurements and may lead to enhanced value of 18F-NaF aortic valve PET in guiding clinical management and risk-stratification of patients with aortic stenosis.