Direct reconstruction of kinetic parameter images from dynamic PET data

摘要: It is often necessary to estimate the parameters of a compartmental model from PET image data. These kinetic are important because they quantify physiological processes. Existing methods for computing parametric images work by first reconstructing sequence images, and then estimating each voxel location in images. We propose novel iterative tomographic reconstruction algorithm directly MAP parameter dynamic sinogram This process estimates vector at using explicit models measurement noise, temporal tracer concentration, spatial variation. Experimental simulations two tissue compartment show that our method can substantially reduce estimation error.