Fast Voxel-Wise Kinetic Modeling in Dynamic PET using a Physics-Informed CycleGAN
PositiveArtificial Intelligence
A recent study introduces a physics-informed CycleGAN that enhances dynamic PET imaging by improving tracer kinetic modeling. This advancement simplifies the complex process of estimating arterial input functions, which is crucial for accurate diagnosis and treatment planning in oncology. The promising results show that the model can predict arterial input functions and generate parameter maps that closely match established references, potentially transforming how practitioners approach imaging and treatment.
— via World Pulse Now AI Editorial System
