AR Medical Image Alignment via Optical Code Markers
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing augmented reality (AR) systems face challenges in precisely identifying the position and orientation of objects, and aligning virtual elements with real-world environments, particularly in scientific, engineering, and medical disciplines, where higher resolution is required.
Innovation Solution
The use of optical codes and image visible markers, such as tubing containing contrast media, affixed to the body of a patient, allows for precise alignment of image data sets with actual views of the patient using an AR headset. This involves identifying optical codes with a camera and using image visible markers to scale and align the image data set accurately.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If virtual objects are aligned with physical objects in AR, then the alignment accuracy is improved, but the virtual object's luminescence obscures the underlying physical object
Solution Approach 1:
Optical codes serve as intermediary markers that can be detected by cameras without requiring direct line-of-sight to the physical object. These codes are placed on or near the physical objects and provide reference points for alignment algorithms, enabling accurate registration of virtual and physical spaces while the virtual objects themselves can be positioned precisely without necessarily obscuring the physical objects, as the alignment is established through the intermediary optical codes rather than direct visual overlay
2Measurement precision
If manual alignment methods are used in medical AR, then positioning resolution is improved, but the process becomes time-consuming and cumbersome
Solution Approach 1:
The system employs automatic alignment through optical codes that are automatically detected and processed by the AR application. The alignment process is performed self-service by the system itself without requiring manual intervention from the user. Optical codes are automatically identified in captured images, and the system automatically calculates and applies the necessary transformations to register virtual objects with physical objects, eliminating time-consuming manual alignment procedures while maintaining high positioning resolution
Solution Approach 2:
Manual mechanical alignment procedures are replaced with automated computational algorithms. Instead of requiring physical manipulation and visual adjustment by the user, the system uses image processing, optical code recognition, and computational geometry to automatically determine positions and orientations, calculate transformations, and register virtual objects with physical objects, thereby reducing alignment time while maintaining or improving positioning precision
3Device complexity
If approximate alignment within a few centimeters is used, then the AR system complexity is reduced, but the positioning accuracy is insufficient for scientific and medical disciplines
Solution Approach 1:
Optical codes serve as simple intermediary markers that add minimal complexity to the system while enabling high-precision alignment. These codes can be simple patterns or sequences that are easily captured by standard cameras and processed by image recognition algorithms. The optical codes provide stable, detectable reference points that allow the system to calculate precise positions and orientations, achieving sub-centimeter accuracy without requiring complex hardware or sophisticated algorithms, thus maintaining system simplicity while dramatically improving positioning accuracy for scientific and medical applications
Data Source
AI summary
An AR headset is described to co-localize an image data set with a body of a person. One method can include identifying optical codes with a contrast medium in a tubing on the body of the person using the AR headset. The image data set can be aligned with the body of the person using the fixed position of the image visible marker with the contrast medium in the tubing with respect to the optical code referenced to a representation of the image visible marker. In one configuration, an image data set can be scaled by comparing a measured size of the image visible marker (e.g., gadolinium tubing) in the captured image data set to the known size of the image visible marker. In addition, a center of an optical code may be identified to more accurately align the image data set with a body of a person.


