Patient ROI Tracking With Distortion-Corrected Scan Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing imaging systems struggle to provide a continuous, unobstructed view of a patient's region of interest during medical scans due to occlusion and perspective changes, especially when using regular surveillance cameras mounted on scanning room walls, and fail to distinguish between patient motion and support displacement, complicating motion detection and image reconstruction.
Innovation Solution
An imaging system with static cameras mounted relative to the scanning system, employing fisheye lenses and image post-processing to correct distortions, tracks patient regions of interest and generates distortion-corrected image sequences, ensuring a stable view and accurate motion detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If regular surveillance cameras are mounted on scanning room walls, then the system can capture patient images, but the view is occluded and cannot provide continuous monitoring of the region of interest
Solution Approach 1:
The patent applies the dynamics principle by mounting cameras on the movable patient support table rather than on fixed wall mounts. This allows the cameras to move synchronously with the patient, maintaining a consistent viewing angle and eliminating occlusion problems that occur when the patient moves through the scanning bore. The dynamic positioning ensures continuous, unobstructed views of the region of interest throughout the entire scanning procedure.
2Stability of the object's composition
If cameras are mounted on the patient support, then a stable view is provided during movement, but the view is obstructed by the patient's abdomen for heavy patients
Solution Approach 1:
The patent applies segmentation by using multiple cameras positioned at different locations on the patient support table. Instead of relying on a single camera that would be blocked by the patient's body, multiple cameras capture images from different angles and positions. This segmented approach ensures that at least one camera maintains an unobstructed view of the region of interest, even when the patient's anatomy blocks other viewing angles.
3Measurement precision
If a camera is placed on the scanning system facing the patient, then images can be captured, but apparent motion due to patient support displacement cannot be separated from true patient motion
Solution Approach 1:
The patent applies mechanics substitution by replacing complex mechanical motion analysis with computational image processing. Instead of using multiple physical sensors or complex mechanical reference systems to distinguish between support displacement and patient motion, the system uses image processing algorithms to analyze the captured images. These algorithms can separate the apparent motion caused by support displacement from true patient motion based on patterns and characteristics in the image data, achieving accurate motion detection without additional mechanical complexity.
4Area of stationary object
If wide field of view cameras are used to monitor the patient, then more of the patient can be seen, but perspective distortion and image distortion increase
Solution Approach 1:
The patent applies the intermediary principle by introducing specialized image processing algorithms as an intermediary between the wide field of view camera and the final displayed image. These algorithms act as a mediator that corrects the perspective distortion and image distortion introduced by the wide angle lens. By processing the raw images through these computational intermediaries, the system maintains the broad field of view coverage while eliminating the unwanted distortion effects, achieving both wide coverage and high image quality.
Data Source
AI summary
An imaging system is for capturing images of a patient during a medical scan. Image post-processing corrects distortions within the captured images from one or more static cameras, and at least one region of interest of the patient is tracked during patient support displacement. A respective distortion-corrected image sequence of the regions of interest is then output.


