[en] This paper systematically evaluates a pharmacokinetic compartmental model for identifying tumor hypoxia using dynamic positron emission tomography (PET) imaging with ¹⁸F-fluoromisonidazole (FMISO). A generic irreversible one-plasma two-tissue compartmental model was used. A dynamic PET image dataset was simulated with three tumor regions-normoxic, hypoxic and necrotic-embedded in a normal-tissue background, and with an image-based arterial input function. Each voxelized tissue's time activity curve (TAC) was simulated with typical values of kinetic parameters, as deduced from FMISO-PET data from nine head-and-neck cancer patients. The dynamic dataset was first produced without any statistical noise to ensure that correct kinetic parameters were reproducible. Next, to investigate the stability of kinetic parameter estimation in the presence of noise, 1000 noisy samples of the dynamic dataset were generated, from which 1000 noisy estimates of kinetic parameters were calculated and used to estimate the sample mean and covariance matrix. It is found that a more peaked input function gave less variation in various kinetic parameters, and the variation of kinetic parameters could also be reduced by two region-of-interest averaging techniques. To further investigate how bias in the arterial input function affected the kinetic parameter estimation, a shift error was introduced in the peak amplitude and peak location of the input TAC, and the bias of various kinetic parameters calculated. In summary, mathematical phantom studies have been used to determine the statistical accuracy and precision of model-based kinetic analysis, which helps to validate this analysis and provides guidance in planning clinical dynamic FMISO-PET studies.

[en] We evaluate a fully data-driven method for the combined recovery and motion blur correction of small solitary pulmonary nodules (SPNs) in F-18 fluorodeoxyglucose (FDG) positron emission tomography (PET)/computed tomography (CT). The SPN was segmented in the low-dose CT using a variable Hounsfield threshold and morphological constraints. The combined effect of limited spatial resolution and motion blur in the SPN's PET image was then modelled by an effective Gaussian point-spread function (psf). Both isotropic and non-isotropic psfs were used. To validate the method, PET/CT measurements of the NEMA/IEC spheres phantom were performed. The method was applied to 50 unselected SPNs ≤30 mm from routine patient care. Recovery of standardised uptake value (SUV) in the phantom image was significantly improved by combined recovery and motion blur correction compared with recovery-only correction, particularly with the non-isotropic model (residual average error 10%). In the patient images, automated segmentation and fit of the effective psf worked properly in all cases. Volume-equivalent diameter ranged from 4.9 to 27.8 mm. Uncorrected maximum SUV ranged from 0.9 to 13.3. Compared with recovery-only correction, combined correction with the non-isotropic model resulted in a 'relevant' (≥30%) SUV increase in 47 SPNs (94%). Correction of both recovery and motion blur is mandatory for accurate SUV quantification of SPNs. (orig.)