New analytical methods in kinetic modeling and reconstruction tomography of dynamic cardiac PET data will be investigated. Compensation for the compound motion (beating and respiratory) of the heart; model selection and parameter identifiability; and tomographic statistical efficiency are major areas of emphasis. Heart motion is the limiting factor determining quantification of tissue tracer concentration in cardiac PET. A detailed study of the bias introduced in the estimation of cardiac tissue tracer concentration and the improvement attributable to new methods of compensation for heart motion is planned. Simulation studies and description length analyses are planned to investigate quantitative and qualitative differences among kinetic models with varying levels of physiologic detail. Particular attention will be given to the PET measurement environment and statistical properties of the measured data. Results will be used to select models for specific clinical objectives and to reconcile models used for kinetic cardiac PET and isolated perfused heart studies. Statistical efficiency of tomographic reconstruction algorithms and the statistically efficient estimation of functionals applied to volume tomographic datasets will be pursued with the objective of improved information extraction from existing PET technology and the reduction of patient dose and discomfort for PET instrumentation of the future.
Showing the most recent 10 out of 55 publications
