A 4-D Iterative HYPR Denoising Operator Improves PET Image Quality

There is an increasing demand for high spatial and/or temporal resolution dynamic PET images in research and clinical settings. Such often have a low number of acquired counts per voxel, leading to poor signal-to-noise ratio, thus hampering quantitative accuracy precision image features. This can be obviated by bias-free postprocessing denoising algorithm improve while preserving feature accuracy. Highly constrained backprojection (HYPR) that offers substantial using 3-D composite image—usually weighted sum over all frames. However, HYPR still introduces bias frames where the contrast differs from composite, limited noise level. In this work, we extend operator iterative 4-D minimize potential mismatching between The initial generated regional averages instead sums level denoising. Through phantom, simulation, human studies, demonstrate (IHYPR4D) yields improved compared traditional HYPR.