Mixed Reality Patient Alignment via 3D Registration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional patient alignment methods for medical devices, such as linear accelerators and CT simulators, face challenges due to non-rigid patient anatomy, field of view limitations, camera obstructions, and high costs, particularly with surface-guided radiation therapy (SGRT), which restricts accurate alignment and is cumbersome and expensive.
Innovation Solution
The use of mixed reality (MR) applications with head-mounted displays for inside-out tracking, allowing dynamic patient alignment by registering a 3D representation of the patient with a medical device, leveraging marker-based and marker-free tracking to overcome field of view limitations and provide real-time alignment feedback, enabling non-rigid patient positioning and reducing costs.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If surface-guided radiation therapy (SGRT) is used for patient alignment, then alignment accuracy can be achieved, but the system becomes cumbersome and expensive
Solution Approach 1:
The patent creates a 3D digital copy of the patient's anatomy from CT scan data and overlays it with the treatment device coordinate system. This virtual 3D representation allows alignment to be performed through image processing and computational methods rather than complex physical measurement systems, reducing device complexity while maintaining alignment accuracy.
Solution Approach 2:
The patent replaces mechanical surface-guided alignment systems with a computational approach using 3D image data and computer vision algorithms. By substituting mechanical measurement and physical alignment tools with digital modeling and automated registration algorithms, the system achieves alignment functionality with reduced mechanical complexity.
2Measurement precision
If conventional alignment methods are used, then patient positioning can be performed, but field of view limitations and camera obstructions restrict accurate alignment
Solution Approach 1:
The patent transitions from 2D camera-based alignment to 3D volumetric alignment by utilizing CT scan data that captures the entire patient anatomy in three dimensions. This dimensional transition allows the system to track and align features throughout the entire 3D space without being constrained by the limited field of view of conventional 2D cameras or affected by obstructions.
Solution Approach 2:
The patent creates a universal 3D representation of the patient's anatomy that can be used for multiple alignment purposes simultaneously. The same 3D model serves as the reference for positioning the treatment device, tracking patient movement, and verifying alignment, eliminating the need for multiple specialized alignment systems with different field of view limitations.
3Productivity
If mixed reality visualization is implemented, then real-time alignment feedback is provided, but additional hardware and processing requirements are introduced
Solution Approach 1:
The patent merges the 3D patient model, treatment device representation, and alignment information into a single integrated mixed reality visualization displayed through head-mounted displays. By combining multiple alignment functions and visual information streams into one unified interface, the system provides real-time feedback while consolidating hardware requirements rather than multiplying them.
Solution Approach 2:
The patent introduces head-mounted displays as an intermediary between the alignment system and the operator. This intermediary device delivers the complex alignment information and 3D visualizations directly to the user's field of view, improving alignment efficiency by providing continuous real-time feedback without requiring additional external monitoring equipment or complex multi-device integration.
Data Source
AI summary
According to at least one aspect, a series of images of a scene are obtained, such as by a head mounted display. The scene includes a medical device and a tracking device, and the series of images including at least the tracking device. Relative pose information indicative of a relative spatial relationship between the tracking device and the medical device is accessed, the 3D representation of the patient is registered with the medical device using the series of images of the scene and the relative pose information, and a mixed reality visualization of the 3D representation of the patient and the medical device generating, based on results of the registering. The mixed reality visualization can further include a visual indication indicative of an alignment of the patient with the 3D representation.