Parametric Methods for Dynamic 11 C-Phenytoin PET Studies.

In this study, the performance of various methods for generating quantitative parametric images of dynamic 11 C-phenytoin PET studies was evaluated. Methods: Double-baseline 60-min dynamic 11 C-phenytoin PET studies, including online arterial sampling, were acquired for 6 healthy subjects. Parametric images were generated using Logan plot analysis, a basis function method, and spectral analysis. Parametric distribution volume (VT ) and influx rate ( K 1 ) were compared with those obtained from nonlinear regression analysis of time-activity curves. In addition, global and regional test-retest (TRT) variability was determined for parametric K 1 and VT values. Results: Biases in VT observed with all parametric methods were less than 5%. For K 1 , spectral analysis showed a negative bias of 16%. The mean TRT variabilities of VT and K 1 were less than 10% for all methods. Shortening the scan duration to 45 min provided similar VT and K 1 with comparable TRT performance compared with 60-min data. Conclusion: Among the various parametric methods tested, the basis function method provided parametric VT and K 1 values with the least bias compared with nonlinear regression data and showed TRT variabilities lower than 5%, also for smaller volume-of-interest sizes (i.e., higher noise levels) and shorter scan duration.