SDED-Reg: A structural discriminability-enhanced decoupled model for unsupervised multi-modal mouse brain optical microscopy image registration

Key Points

• Image registration improves under a new decoupled model approach, leading to better alignment of multi-modal data.
• The evaluation shows a significant improvement in structural discriminability, enhancing optical microscopy accuracy.
• Analysis utilizes an unsupervised approach for effective registration of mouse brain images from varied modalities.
• These findings indicate potential advances for research in neurological imaging, supporting future applications in brain studies.