3D-2D Image Registration Using Segmented MRI Projections
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Intraoperative image-guided clinical interventions face challenges due to the time-sensitive nature of target localization in 2D images, exacerbated by mismatches in image intensities and anatomical details between preoperative and intraoperative imaging modalities, which complicates accurate 3D-2D registration.
Innovation Solution
A method for 3D-2D registration that segments structures of interest in preoperative 3D MRI images, generates simulated projections, and aligns them with 2D images, using a non-transitory computer readable medium to display registered 3D data onto 2D images, overcoming mismatches by focusing on relevant anatomical regions and using gradient orientation similarity metrics and CMA-ES optimizers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If 3D-2D registration is performed using preoperative imaging data and intraoperative images, then target localization accuracy is improved, but image intensity and anatomical detail mismatches between different imaging modalities worsen registration performance
Solution Approach 1:
The patent segments the 3D preoperative image into multiple anatomical structures or regions of interest before performing registration. This segmentation allows the system to focus on specific anatomical features that are most relevant for localization, reducing the impact of mismatches in other regions. The segmented structures can be independently registered and transformed to the 2D intraoperative space, improving overall registration reliability.
Solution Approach 2:
The patent extracts specific anatomical features or landmarks from the 3D preoperative image that are most suitable for registration with the 2D intraoperative image. By selecting and extracting only the most relevant anatomical details rather than using the entire 3D volume, the system reduces mismatches caused by modality differences and improves registration performance.
2Loss of information
If comprehensive 3D image data is registered to 2D images, then anatomical context is improved, but processing time and computational complexity worsen
Solution Approach 1:
The patent divides the computationally intensive 3D image processing into smaller segmented regions that can be processed independently and in parallel. This segmentation reduces the overall computational complexity and processing time while preserving the essential anatomical context needed for accurate localization.
Solution Approach 2:
The patent extracts only the essential anatomical features and contextual information needed for registration, rather than processing the complete 3D image data. This extraction approach maintains sufficient anatomical context for accurate localization while significantly reducing computational burden and processing time.
3Loss of information
If manual interpretation of intraoperative images is used for target localization, then clinical decision support is provided, but clinician stress and time constraints worsen workflow efficiency
Solution Approach 1:
The patent implements an automated system that performs 3D-2D registration and target localization without requiring manual clinician interpretation. The system automatically processes the imaging data, performs the registration transformations, and provides localization results, thereby reducing clinician stress and improving workflow efficiency while maintaining accurate clinical decision support.
Data Source
AI summary
An embodiment in accordance with the present invention provides a technique for localizing structures of interest in projection images (e.g., x-ray projection radiographs or fluoroscopy) based on structures defined in a preoperative 3D image (e.g., MR or CT). Applications include, but are not limited to, spinal interventions. The present invention achieves 3D-2D image registration (and particularly allowing use with a preoperative MR image) by segmenting the structures of interest in the preoperative 3D image and generating a simulated projection of the segmented structures to be aligned with the 2D projection image. Other applications include various clinical scenarios involving 3D-2D image registration, such as image-guided cranial neurosurgery, orthopedic surgery, biopsy, and radiation therapy.