Erlandsson, K;
Liljeroth, M;
Atkinson, D;
Arridge, S;
Ourselin, S;
Hutton, BF;
(2016)
Improved Parameter-Estimation With MRI-Constrained PET Kinetic Modeling: A Simulation Study.
IEEE Transactions on Nuclear Science
, 63
(5)
pp. 2464-2470.
10.1109/TNS.2015.2507444.
Preview |
Text
07422171.pdf - Published Version Download (1MB) | Preview |
Abstract
Kinetic analysis can be applied both to dynamic PET and dynamic contrast enhanced (DCE) MRI data. We have investigated the potential of MRI-constrained PET kinetic modeling using simulated [ ^{18}{\rm F} ]2-FDG data for skeletal muscle. The volume of distribution, {V_{\rm e}} , for the extra-vascular extra-cellular space (EES) is the link between the two models: It can be estimated by DCE-MRI, and then used to reduce the number of parameters to estimate in the PET model. We used a 3 tissue-compartment model with 5 rate constants (3TC5k), in order to distinguish between EES and the intra-cellular space (ICS). Time-activity curves were generated by simulation using the 3TC5k model for 3 different {V_{\rm e}} values under basal and insulin stimulated conditions. Noise was added and the data were fitted with the 2TC3k model and with the 3TC5k model with and without {V_{\rm e}} constraint. One hundred noise-realisations were generated at 4 different noise-levels. The results showed reductions in bias and variance with {V_{\rm e}} constraint in the 3TC5k model. We calculated the parameter {k_3}^{\prime \prime } , representing the combined effect of glucose transport across the cellular membrane and phosphorylation, as an extra outcome measure. For {k_3}^{\prime \prime } , the average coefficient of variation was reduced from 52% to 9.7%, while for {k}_{3} in the standard 2TC3k model it was 3.4%. The accuracy of the parameters estimated with our- new modeling approach depends on the accuracy of the assumed {V_{\rm e}} value. In conclusion, we have shown that, by utilising information that could be obtained from DCE-MRI in the kinetic analysis of [ ^{18}{\rm F} ]2-FDG-PET data, it is in principle possible to obtain better parameter estimates with a more complex model, which may provide additional information as compared to the standard model.
Archive Staff Only
 |
View Item |
