Adding New Similarity Metrics for DIPY’s Image Registration Frameworks

This proposal aims to extend DIPY's image registration framework by implementing new similarity metrics, with a particular focus on Mutual Information (MI) for nonrigid registration. While MI is currently available in DIPY for affine registration, support for nonrigid registration is not yet implemented despite its importance for multimodal medical imaging. This project will design, implement, and integrate an efficient MI-based similarity metric within DIPY's existing registration pipeline, together with appropriate validation and testing.