Automatic Fluorescence Artifact Removal in OCT-NIRAF Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current imaging technologies, such as MMOCT, face challenges in automatically detecting and correcting fluorescence signal artifacts in catheter-based multimodality OCT-NIRAF images, which can lead to misleading clinical interpretations due to artifacts from sources like stent struts, guide wires, and irregular vessel shapes, requiring manual review and increasing user fatigue.
Innovation Solution
A method involving automatic thresholding and unsupervised machine learning classification, specifically using the DBSCAN algorithm, to detect and correct fluorescence artifacts in MMOCT images by analyzing standard deviation and perpendicular distance, replacing noisy values with corresponding values from adjacent frames.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual review of fluorescence artifacts is performed, then diagnostic accuracy is improved, but user fatigue increases and efficiency decreases
Solution Approach 1:
The system performs automatic artifact detection and removal without requiring manual user intervention. The fluorescence artifact removal module autonomously identifies and corrects artifacts in OCT-NIRAF images, eliminating the need for users to manually review and correct each artifact, thus reducing user fatigue while maintaining diagnostic accuracy
Solution Approach 2:
The patent replaces the manual mechanical review process with an automated computational system. The fluorescence artifact removal module uses algorithmic processing to detect and remove artifacts, substituting the manual visual inspection and correction process with an automated digital system that maintains accuracy while improving efficiency
2Manufacturing precision
If fluorescence artifact removal is performed manually, then image quality is improved, but processing time increases
Solution Approach 1:
The system automatically detects and removes fluorescence artifacts without requiring manual user intervention for each image. The fluorescence artifact removal module autonomously processes images in real-time during the pullback procedure, eliminating the time-consuming manual review and correction process while maintaining high image quality
Solution Approach 2:
The patent performs artifact removal as a preliminary step in the image processing workflow, before final image display and interpretation. The fluorescence artifact removal module proactively identifies and corrects artifacts in advance, preventing them from interfering with subsequent diagnostic analysis and eliminating the need for time-consuming post-processing manual correction
3Productivity
If automatic artifact detection is implemented, then efficiency is improved, but false positive detection may occur
Solution Approach 1:
The patent employs adaptive thresholding that dynamically adjusts detection parameters based on the specific characteristics of each image and pullback procedure. The fluorescence artifact removal module modifies its detection sensitivity and criteria according to the observed signal patterns, allowing it to efficiently detect artifacts while adapting to different imaging conditions to minimize false positives
Solution Approach 2:
The system applies localized detection criteria that consider the specific spatial and signal characteristics of different regions in the image. The fluorescence artifact removal module analyzes artifacts in the context of their local environment, using region-specific thresholds and patterns to distinguish true artifacts from legitimate fluorescence signals, thereby improving detection reliability while maintaining efficiency
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 approach enables efficient and accurate automatic detection and removal of fluorescence artifacts, reducing user intervention and improving image quality by distinguishing true molecular information from artifacts, thereby enhancing diagnostic accuracy in coronary artery disease assessment.
Implementation Method 1
a fiber-based catheter comprising a sheath, an imaging core rotatably mounted in the sheath, a pullback unit, a torque coil, a torque motor, a torque motor driver, a drive shaft, a drive gear, a drive pinion gear, a lead screw, and a lead screw motor
Implementation Method 2
the second detector is configured to detect part of the sample beam transmitted from the fiber coupler
Implementation Method 3
Intravascular fluorescence is a catheter-based molecular imaging technique that uses near-infrared fluorescence to detect artery wall autofluorescence (NIRAF) or artery wall fluorescence generated by molecular agents injected intravenously (NIRF)
Implementation Method 4
The excitation light is guided by a fiber, the FORJ, the FRC, and the distal optics to irradiate the vessel
Implementation Method 5
a fiber coupler configured to couple the sample beam to the circulator and to couple the excitation light to the distal optics
Implementation Method 6
a fluorescence ratio camera configured to detect the fluorescence light emitted by the vessel in a first wavelength range between 660 nm and 740 nm
Data Source
AI summary
Embodiments disclosed herein provide systems, methods and/or computer-readable media for automatically detecting and removing fluorescence artifacts from catheter-based multimodality OCT-NIRAF images. In one embodiment, a process of determining an automatic threshold value (automatic thresholding) is implemented by sorting characteristic parameter values of the NIRAF signal and finding a maximum perpendicular distance between a curve of the sorted values and a straight line from the highest to the lowest sorted value, combined with the use of unsupervised machine learning classification techniques to detect the frame's NIRAF values that correspond to signal artifacts. Once the signal artifacts are detected, the system can filter out the signal artifacts, correct the frames that had artifacts, and produce a more accurate multimodality image.