A technique for extracting physiological parameters and the required input function simultaneously from PET image measurements: theory and simulation study

摘要: Positron emission tomography (PET) is an important tool for enabling quantification of human brain function. However, quantitative studies using tracer kinetic modeling require the measurement time-activity curve in plasma (PTAC) as model input It widely believed that insertion arterial lines and subsequent collection processing biomedical signal sampled from blood are not compatible with practice clinical PET, it invasive exposes personnel to risks associated handling patient radiation dose. Therefore, interest develop practical noninvasive techniques PET. In this paper, a technique proposed extract function together physiological parameters dynamic images alone. The identifiability method tested rigorously by Monte Carlo simulation. results show able quantify all required information obtained two or more regions (ROIs) very different dynamics PET images. There no significant improvement parameter estimation local cerebral metabolic rate glucose (LCMRGlc) if there than three ROIs. can provide reliable LCMRGlc, which our primary study.