Decoupling Multi-Contrast Super-Resolution: Self-Supervised Implicit Re-Representation for Unpaired Cross-Modal Synthesis
The framework decouples Multi-Contrast Super-Resolution into two stages: learning a population anatomical prior using unpaired cross-modal synthesis (uCMS) and reconstructing patient-specific images via implicit re-representation (IrR). This allows high-fidelity reconstruction without needing paired training data, fusing general knowledge with individual subject details.