Optical 3D Pose Tracking for PET Head-Motion Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for correcting head motion in PET brain scanning are either invasive, uncomfortable for patients, or ineffective for long scans, particularly in research studies with neuroligands.
Innovation Solution
The use of optical pose tracking of natural facial features to quickly and accurately determine the pose of the head during PET scanning, without the need for patient preparation, fixtures, or artificial features, and incorporating this data iteratively into image reconstruction for motion correction.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If optical tracking of a target attached to the subject's head is used, then pose can be determined, but the time and effort in securely attaching the target increases complexity and may not always be successful
Solution Approach 1:
The patent extracts the tracking target from the system by using natural facial features themselves as the tracking markers. Instead of attaching artificial targets to the head, the system directly tracks the natural features (eyes, nose, mouth) that already exist on the subject's face, eliminating the attachment step entirely while maintaining pose determination capability
Solution Approach 2:
The system uses the subject's own natural facial features as the tracking targets, making the subject self-sufficient for the tracking process. The facial features serve their dual purpose of identifying the subject and providing pose information without requiring external attachments or preparations
2Measurement precision
If marker-based optical devices for tracking are used, then pose can be tracked, but they are difficult to attach rigidly and require multiple images and high computation
Solution Approach 1:
The patent removes the need for separate marker devices by using natural facial features as the tracking targets. This eliminates the attachment difficulty and reduces the system to using only the subject's own features, simplifying the overall process while maintaining tracking accuracy
Solution Approach 2:
Instead of using physical markers that need to be attached, the system creates a digital copy/tracking model of the natural facial features. The facial features themselves serve as the tracking targets, and the system tracks their positions across images to determine pose, avoiding the need for physical marker attachment and complex processing
3Ease of operation
If co-registration of PET image to reference image is used, then motion parameters can be determined without instrumentation, but it cannot be applied to time scales of minutes and may be inaccurate for evolving tracers
Solution Approach 1:
The patent introduces optical imaging of facial features as an intermediary system between the PET scan and motion measurement. This intermediary captures motion information at high temporal resolution without requiring direct PET-PET co-registration, allowing accurate motion tracking throughout the entire scan duration while maintaining ease of operation
Solution Approach 2:
The system segments the motion correction process into two independent parts: (1) optical tracking of facial features to determine head pose and motion, and (2) PET image reconstruction with motion correction. This segmentation allows each component to be optimized independently, with the optical tracking handling temporal resolution and the PET reconstruction handling image quality, resolving the contradiction between operational ease and measurement precision
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This method effectively corrects blurring and artifacts caused by head motion, achieving improved image quality with reduced patient discomfort and increased accuracy in motion correction.
Implementation Method 1
optical camera to detect and track the 3D position of facial features on the subject's head
Data Source
AI summary
A data-driven method and system using optical tracking of natural facial features for quick and accurate pose determination during scanning using one or more cameras installed within or on a PET-CT gantry. No patient preparation, fixtures, or artificial features are needed. The pose data is analyzed and incorporated iteratively in reconstruction for reconstruction to motion correct the PET data. The present invention thus results in improved correction of blurring and artifacts caused by head motion during the PET scan.