Intrinsic 2D/3D registration based on a hybrid approach: use in the radiosurgical imaging process

Area of Science:

• Medical Imaging
• Radiosurgery
• Image Registration

Background:

• Multimodal image matching is advancing, enabling intrinsic registration using anatomical data.
• Current intrinsic registration techniques are primarily research-based and not widely used clinically.
• Radiosurgery planning for Arteriovenous Malformations (AVMs) relies on multimodality imaging and stereotactic frames.

Purpose of the Study:

• To present a novel method for matching projective (2D) and tomographic (3D) medical imaging modalities.
• To propose an application of this image registration technique for AVM radiosurgery planning.
• To describe the validation process for the proposed image registration approach.

Main Methods:

• Developed a method for matching projective (e.g., radiography, angiography) and tomographic (e.g., MRI, CT) imaging.
• Proposed integrating diagnostic images into the radiosurgery planning process for AVMs.
• Outlined the validation steps for the developed image registration technique.

Main Results:

• The proposed image registration technique was applied to AVM radiosurgery planning.
• Current results do not yet meet the accuracy required to replace conventional stereotactic frame-based planning.
• Analysis indicates potential for future operational use with protocol improvements.

Conclusions:

• The presented image registration method offers a pathway to incorporate diagnostic images into AVM radiosurgery planning.
• Further protocol refinement is necessary to achieve the accuracy required for clinical replacement of conventional techniques.
• The approach demonstrates potential for enhancing AVM treatment planning in the future.