2D-to-3D Image Registration Without Radiopaque Fiducials
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing surgical navigation and robotic guidance systems require the use of radiopaque fiducials for image registration, which limits flexibility, increases procedure time, and exposes patients to unnecessary radiation due to the need for additional imaging, and can obstruct critical anatomy views.
Innovation Solution
A method and system for registering medical images without radiopaque fiducials by using pre-operative 3D images and intra-operative 2D images or synthetically generated 2D images to establish registration, allowing for registration recovery without full 3D scans, thus reducing procedure time and radiation exposure.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If radiopaque fiducials are used for image registration, then registration accuracy is improved, but radiation exposure increases and anatomical views are obstructed
Solution Approach 1:
The patent removes radiopaque fiducials from the registration process entirely. Instead of using physical fiducial markers that require additional imaging, the system uses natural anatomical landmarks and surface geometry from the pre-operative 3D scan to perform registration, thereby eliminating the harmful radiation exposure associated with capturing fiducial-containing images.
Solution Approach 2:
The system creates a digital copy of the patient's anatomy through pre-operative 3D surface scanning. This digital model is then used for registration purposes, replacing the need for physical fiducial markers and subsequent radiographic imaging. The surface geometry itself serves as the registration reference, eliminating radiation exposure while maintaining registration capability.
2Measurement precision
If registration fixtures with fiducials are used, then image registration is achieved, but procedure time increases due to additional imaging requirements
Solution Approach 1:
The patent performs the registration setup in advance by capturing a 3D surface scan of the patient's anatomy during the pre-operative phase. This pre-captured geometric data is then used during surgery for registration without requiring additional time-consuming imaging procedures. The registration references are established before the surgical procedure begins.
Solution Approach 2:
The system extracts and uses only the necessary geometric information from the pre-operative scan for registration purposes, eliminating the need for intraoperative fiducial placement and associated imaging. This extraction of essential registration data from the digital model streamlines the surgical workflow and reduces procedure time.
3Measurement precision
If radiopaque fiducials are embedded in images, then registration can be performed, but usable image space is reduced and critical anatomy may be obstructed
Solution Approach 1:
The patent removes fiducial markers from the imaging field entirely and uses the patient's natural anatomical surface geometry for registration. By extracting registration information from the existing anatomical structures visible in the primary images rather than adding fiducial markers, the system preserves the complete usable image space and maintains unobstructed views of critical anatomy.
Data Source
AI summary
System and method of registering a medical image of a patient in an imaging space to the patient in a physical space preferably without the use of any embedded radiopaque fiducials in medical images is provided. In one way, intra-op 2D medical images are used to register a pre-op unregistered 3D medical image. The 2D medical images are registered based on simultaneous tracking of the tracking markers on the imaging device and on the patient by a tracking device at the time of image capture. The 2D images are matched to corresponding simulated 2D images generated from the pre-op 3D image volume. Thus, registration of a pre-op 3D image to the patient is accomplished without performing another 3D scan of the patient.


