2D-3D Image Registration Verification for Endoscope Targeting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for registering three-dimensional images with two-dimensional fluoroscopic images lack accuracy verification, leading to potential delays in accurately guiding ultrasonic endoscopes to target positions within the body.
Innovation Solution
An image processing device that specifies a target position in a three-dimensional image, performs registration with sequentially acquired two-dimensional images, and evaluates the registration accuracy by deriving a range in which the target position may exist, superimposing this range on the two-dimensional image for confirmation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If registration is performed between three-dimensional image and two-dimensional fluoroscopic image, then the lesion position can be specified in the fluoroscopic image, but the accuracy of registration cannot be verified
Solution Approach 1:
The patent introduces a pseudo two-dimensional image as an intermediary representation of the three-dimensional image. This pseudo 2D image serves as a mediator that can be directly compared with the actual fluoroscopic image to verify registration accuracy. The pseudo 2D image is generated by projecting the 3D image data onto a 2D plane, creating a reference that bridges the gap between 3D and 2D spaces for accuracy verification purposes.
Solution Approach 2:
The patent performs registration not only between the actual fluoroscopic image and the three-dimensional image, but also generates and compares a pseudo two-dimensional image with the fluoroscopic image. This partial action of creating an additional reference image allows for verification of registration accuracy without requiring a complete redesign of the imaging system.
2Measurement precision
If fluoroscopic imaging is performed continuously to guide ultrasonic endoscope to target position, then real-time positioning is achieved, but the procedure takes time due to insufficient registration accuracy
Solution Approach 1:
The patent implements a feedback mechanism by comparing the pseudo two-dimensional image with the actual fluoroscopic image. This comparison provides feedback on the registration accuracy, allowing the system to verify whether the three-dimensional image is correctly aligned with the two-dimensional fluoroscopic image. This feedback loop enables accurate target positioning without requiring prolonged fluoroscopic imaging.
Solution Approach 2:
The patent performs registration and generates the pseudo two-dimensional image in advance before the actual treatment procedure. By preparing the registration and verification data beforehand, the system eliminates the need for time-consuming adjustments and verifications during the procedure, thereby reducing overall procedural time while maintaining accuracy.
3Reliability
If three-dimensional image is acquired before treatment and registered with fluoroscopic image, then lesion position can be specified, but registration accuracy may be insufficient
Solution Approach 1:
The patent introduces a pseudo two-dimensional image as an intermediary representation of the three-dimensional image. This pseudo 2D image serves as a mediator that can be directly compared with the actual fluoroscopic image to verify registration accuracy. The pseudo 2D image is generated by projecting the 3D image data onto a 2D plane, creating a reference that bridges the gap between 3D and 2D spaces for accuracy verification purposes.
Solution Approach 2:
The patent changes the dimensional parameter of the three-dimensional image by generating a two-dimensional projection (pseudo 2D image) from the 3D data. This parameter change allows for direct comparison with the fluoroscopic image and enables accuracy verification. The system transforms the 3D image into a 2D representation while preserving the essential spatial relationships needed for registration verification.
Data Source
AI summary
A processor specifies a target position in a three-dimensional image acquired by imaging a subject before treatment, performs registration between two-dimensional images sequentially acquired for the subject under the treatment and the three-dimensional image, derives an evaluation result representing a reliability of the registration at least at the target position of the three-dimensional image, derives a range in which the target position possibly exists in the two-dimensional image based on a result of the registration and the evaluation result, and displays the two-dimensional image on which the range is superimposed.


