Electronic Thesis and Dissertation Repository


Doctor of Philosophy


Medical Biophysics


Ting-Yim Lee


Functional imaging of 18F-Fluorocholine PET holds promise in the detection of dominant prostatic lesions. Quantitative parameters from PET-CT Perfusion may be capable of measuring choline kinase activity, which could assist in identification of the dominant prostatic lesion for more accurate targeting of biopsies and radiation dose escalation.

The objectives of this thesis are: 1) investigate the feasibility of using venous TACs in quantitative graphical analysis, and 2) develop and test a quantitative PET-CT Perfusion imaging technique that shows promise for identifying dominant prostatic lesions.

Chapter 2 describes the effect of venous dispersion on distribution volume measurements with the Logan Plot. The dispersion of venous PET curves was simulated based on the arterio-venous transit time spectrum measured in a perfusion CT study of the human forearm. The analysis showed good agreement between distribution volume measurements produced by the arterial and venous TACs.

Chapter 3 details the mathematical implementation of a linearized solution of the 3-Compartment kinetic model for hybrid PET-CT Perfusion imaging. A noise simulation determined the effect of incorporating CT perfusion parameters into the PET model on the accuracy and variability of measurements of the choline kinase activity. Results indicated that inclusion of CT perfusion parameters known a priori can significantly improve the accuracy and variability of imaging parameters measured with PET.

Chapter 4 presents the implementation of PET-CT Perfusion imaging in a xenograft mouse model of human prostate cancer. Image-derived arterial TACs from the left ventricle were corrected for partial volume and spillover effects and validated by comparing to blood sampled curves. The PET-CT Perfusion imaging technique produced parametric maps of the choline kinase activity, k3. The results showed that the partial volume and spillover corrected arterial TACs agreed well with the blood sampled curves, and that k3max was significantly correlated with tumor volume, while SUV was not.

In summary, this thesis establishes a solid foundation for future clinical research into 18F-fluorocholine PET imaging for the identification of dominant prostatic lesions. Quantitative PET-CT Perfusion imaging shows promise for assisting targeting of biopsy and radiation dose escalation of prostate cancer.