Three Dimensional MR Image Synthesis with Progressive Generative Adversarial Networks
This work aims to improve the quality of 3D MRI synthesis for medical imaging applications by addressing common artifacts and resolution issues.
This paper addresses the challenge of 3D MRI synthesis, where existing methods either suffer from discontinuity artifacts (cross-sectional) or loss of spatial resolution (volumetric). The authors propose a novel model that progressively reconstructs the target volume by performing simpler synthesis tasks across individual orientations, aiming to mitigate the limitations of both prior approaches.
Mainstream deep models for three-dimensional MRI synthesis are either cross-sectional or volumetric depending on the input. Cross-sectional models can decrease the model complexity, but they may lead to discontinuity artifacts. On the other hand, volumetric models can alleviate the discontinuity artifacts, but they might suffer from loss of spatial resolution due to increased model complexity coupled with scarce training data. To mitigate the limitations of both approaches, we propose a novel model that progressively recovers the target volume via simpler synthesis tasks across individual orientations.