Virtual Fiducial Mapping for Markerless Motion Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing medical imaging techniques face challenges with subject motion during data acquisition, leading to image corruption or artifacts, particularly when using fiducial markers that require application to the subject's surface, and existing markerless motion tracking systems are limited in flexibility and robustness.
Innovation Solution
Implementing a system that uses an image marking neural network to generate virtual fiducial markers within the camera coordinate system, which are then converted to the medical imaging system's coordinate system, allowing for robust motion correction and field of view adjustment without physical markers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If physical fiducial markers are applied to the subject's surface for motion tracking, then motion correction capability is improved, but subject comfort and ease of operation deteriorate due to marker application requirements
Solution Approach 1:
The patent creates virtual fiducial markers by generating synthetic images of markers and superimposing them onto the subject's anatomy using image registration techniques. These virtual markers are then tracked through the imaging sequence without requiring any physical markers to be applied to the subject, thus maintaining motion correction capability while eliminating the discomfort and complexity of physical marker application
Solution Approach 2:
The patent replaces the mechanical system of physical fiducial markers with a computational approach using virtual markers generated through image processing and registration. This substitution eliminates the need for physical objects while maintaining the functional capability of tracking subject motion through the imaging sequence
2Ease of operation
If markerless motion tracking is implemented, then ease of operation is improved, but tracking reliability and robustness deteriorate
Solution Approach 1:
The patent introduces virtual fiducial markers as an intermediary element that bridges the gap between markerless operation and reliable tracking. These virtual markers serve as stable reference points for motion estimation without requiring physical application to the subject, thus maintaining both ease of operation and tracking reliability
Solution Approach 2:
The patent creates synthetic representations of fiducial markers that can be reliably tracked through the imaging sequence. These copied virtual markers provide stable reference points for motion correction without requiring physical objects, thereby maintaining tracking reliability while preserving the benefits of markerless operation
3Ease of operation
If virtual fiducial markers are used instead of physical markers, then ease of operation is improved by eliminating marker application, but measurement precision may deteriorate
Solution Approach 1:
The patent replaces physical markers with computationally-generated virtual markers that are superimposed onto the subject's anatomy through image registration. This substitution maintains measurement precision by using the same coordinate systems and registration techniques that ensure accurate spatial localization, while eliminating the need for physical marker application
Solution Approach 2:
The patent transforms the physical marker concept into a virtual marker representation by changing the state from physical object to digital image data. This parameter change allows the markers to be precisely localized through image registration algorithms while maintaining ease of operation by eliminating physical application requirements
Data Source
AI summary
Disclosed herein is a medical system (100, 300, 500, 600) comprising: a medical imaging system (102, 302) configured to acquire medical imaging data (136) descriptive of a subject (110); a camera system (114) configured to acquire a subject image (138) of the subject; a memory (126) storing machine executable instructions (130), medical imaging system commands (134), and a coordinate system mapping, and an image marking neural network (132). Execution of the machine executable instruction by a computational system (120) causes the computational system to: acquire (200) the medical imaging data by controlling the medical imaging system with the medical imaging system commands; repeatedly (202) control the camera system to acquire the subject image during acquisition of the medical imaging data; repeatedly (204) receive camera system coordinates (142) of the virtual fiducial markers by inputting the subject image into the image marking neural network; and repeatedly (206) provide imaging system coordinates of the set of virtual fiducial markers by repeatedly converting the camera system coordinates of the virtual fiducial markers to the provided imaging system coordinates of the virtual fiducial markers using the coordinate system mapping.


