Hyperspectral Endoscopic Imaging With Motion Artifact Collation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing medical imaging technologies fail to address the issue of motion artifacts in hyperspectral imaging, which occur due to movement of the spectral camera and the area under investigation relative to each other during the recording process, leading to reduced image quality or unusable images.
Innovation Solution
A method and device that includes a collation process to detect and correct motion artifacts by using a reference, such as a white-light image, to align the spectral camera with the area under investigation, employing sensors and tracking systems to monitor and adjust the spectral image in real-time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If time-shifted spatial scanning is used to acquire hyperspectral images in multiple line steps, then spectral resolution and imaging capability are improved, but motion artifacts occur due to sequential acquisition causing parts of the area to be scanned multiple times or skipped
Solution Approach 1:
The patent applies preliminary action by acquiring a reference image (such as a white light image or another spectral image) before the main spectral imaging acquisition. This reference image serves as a baseline for detecting and correcting motion artifacts in the spectral images, allowing the system to compensate for movements that occur during the sequential line-by-line scanning process.
Solution Approach 2:
The patent implements feedback by continuously comparing sequentially acquired spectral images or line steps against the reference image to detect motion artifacts. The detected motion information is then used to correct the spectral images, creating a closed-loop system that actively compensates for camera or tissue movement during acquisition.
2Loss of information
If sequential line-by-line scanning is performed to capture spectral information, then comprehensive spectral data is obtained, but uneven exposures occur due to movement causing incorrect intensities of spectral information
Solution Approach 1:
The reference image is acquired beforehand to establish the correct exposure levels and spectral intensities for each region. This preliminary capture provides a template against which subsequent spectral images are compared, allowing the system to identify and correct exposure inconsistencies caused by motion during the scanning process.
Solution Approach 2:
The patent changes the parameter being measured by comparing intensity values across different time points (reference image vs. spectral images). By detecting changes in spectral intensity patterns that correspond to motion, the system can adjust and normalize the spectral data to compensate for uneven exposures caused by camera or tissue movement.
3Reliability
If motion detection and correction processes are implemented, then image quality is improved, but system complexity and processing requirements increase
Solution Approach 1:
The patent uses a simplified copy approach by acquiring a reference image that replicates the imaging conditions without requiring complex real-time motion tracking hardware. This reference copy serves as a basis for detecting motion artifacts in the spectral images through comparative analysis, reducing the need for additional complex sensors or tracking systems.
Solution Approach 2:
The patent replaces complex mechanical motion tracking systems with a computational approach. Instead of using additional mechanical sensors or tracking devices to monitor camera or tissue movement, the system uses image processing algorithms to detect motion artifacts by comparing the spectral images against the reference image, substituting mechanical complexity with computational analysis.
Data Source
Figure 1
Figure 2~3
Figure 4~5
AI summary
The invention relates to a method for medical imaging in which, in a single acquisition process, at least one multi- and/or hyperspectral spectral image of an examination area is acquired by an endoscopic and/or exoscopic multi- and/or hyperspectral spectral camera. It is proposed that the method comprises at least one collation process in which motion artifacts in the spectral image, which occur due to movement of the spectral camera and the examination area relative to each other during the acquisition process, are at least partially reduced and/or avoided by taking into account at least one reference of the spectral camera and/or the examination area, wherein, on the basis of the reference, movement of the spectral camera relative to the examination area is indicated and/or detected.