Fast and Robust Symmetric Image Registration Based on Distances Combining Intensity and Spatial Information

Area of Science:

• Medical Imaging
• Computer Vision
• Image Processing

Background:

• Intensity-based image registration is crucial for aligning images, relying heavily on similarity measures.
• The accuracy and robustness of registration depend significantly on the properties of these similarity measures.

Purpose of the Study:

• To propose a novel symmetric, intensity interpolation-free affine registration framework.
• To evaluate the framework's performance using both synthetic and real-world 2D and 3D medical image data.

Main Methods:

• Developed a registration framework integrating intensity and spatial information.
• Tested the framework within a standard gradient-based registration pipeline using the Insight Segmentation and Registration Toolkit (ITK).

Main Results:

• The proposed framework demonstrated superior robustness and accuracy compared to commonly used similarity measures.
• Effective performance was validated on synthetic data with noise and real biomedical/medical imaging applications.
• Low computational cost was empirically confirmed, indicating practical applicability.

Conclusions:

• The novel registration framework offers improved accuracy and robustness for medical image registration.
• Its efficiency and effectiveness make it a valuable tool for 2D and 3D image analysis.
• The availability of source code facilitates further research and application.