CycleINR: A Flexible and Consistent Approach for Arbitrary-Scale Volumetric Super-Resolution of Medical Imaging Data

CycleINR, a novel framework for 3D medical image volumetric super-resolution, leverages the continuity of implicit neural representation (INR) to achieve flexible and consistent high-resolution image generation without specific training for upsampling ratios.