Pre-computed DRR Knowledge Base for Real-time Image Registration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods for registering pre-operative and intra-operative images are computationally intensive, taking several minutes to complete, which is not suitable for real-time applications in minimally invasive therapeutic interventions, and compromise on robustness when reducing computation time.
Innovation Solution
Pre-computing Digitally Reconstructed Radiographs (DRRs) for various poses and storing them in a knowledge base, allowing for efficient retrieval and registration of pre-operative and intra-operative images in real-time by matching signatures from both sets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If conventional registration methods are used to register pre-operative and intra-operative images, then registration accuracy is maintained, but registration time becomes too long (several minutes) for real-time applications
Solution Approach 1:
The patent pre-computes and stores a knowledge base of DRRs (Digitally Reconstructed Radiographs) for various poses before the actual registration process. This preliminary computation allows the system to quickly retrieve and match images during real-time procedures, reducing registration time from several minutes to near real-time speeds while maintaining accuracy.
2Speed
If computation time is reduced by randomly sampling DRRs, then registration speed improves, but robustness of results is compromised due to less information available to the optimizer
Solution Approach 1:
Instead of randomly sampling during the actual registration process, the system pre-computes and stores a comprehensive knowledge base of DRRs for various poses. This allows the system to retrieve relevant pre-computed images for matching, maintaining both speed and robustness by having full information available without requiring time-consuming optimization iterations.
3Measurement precision
If three dimensional images are used to provide high resolution anatomical detail, then image quality improves, but acquisition time and intrusiveness to the physician increase significantly
Solution Approach 1:
The patent separates the imaging modalities into two segments: pre-operative 3D imaging (CT or MRI) for acquiring high-resolution anatomical details, and intra-operative 2D imaging (X-ray or fluoroscopy) for real-time procedural guidance. This segmentation allows each modality to serve its optimal purpose - detailed anatomy from 3D pre-op images and quick, less intrusive 2D images during the procedure.
Solution Approach 2:
The system performs the time-consuming 3D image acquisition and DRR generation before the actual surgical procedure. By pre-computing the knowledge base of DRRs from pre-operative 3D images, the system eliminates the need to acquire 3D images in real-time during the procedure, maintaining high resolution while reducing intra-operative time and intrusiveness.
Data Source
AI summary
A system and method for registering pre-operative images of an object with an intra-operative image of the object is disclosed. Prior to an operative procedure, Digitally Reconstructed Radiographs (DRRs) are generated for the pre-operative images of each individual patient. Signatures are extracted from the DRRs. The signatures are stored in a knowledge base. During the operative procedure, a signature is extracted from the intra-operative image. The intra-operative signature is compared to the stored pre-operative signatures. A pre-operative image having a best signature match to the intra-operative signature is retrieved. The retrieved pre-operative image is registered with the intra-operative image.