Machine Learning Fluorescence Imaging Background Suppression

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Current fluorescence imaging in surgical procedures is compromised by undesirable background signals due to impractical timing of fluorescent dye administration, making it difficult for surgeons to accurately identify structures of interest, such as the biliary tree, during surgeries.

Innovation Solution

The use of machine-learning models to enhance fluorescence medical images by identifying background areas and suppressing unwanted signals relative to areas of interest, based on both visible-light and fluorescence images, allowing for intelligent identification and enhancement of relevant anatomical structures.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If fluorescent dye is administered 15-60 minutes before surgery, then the dye is readily available and easy to administer, but background signals from hepatic tissue increase making it difficult to identify structures of interest

Engineering Contradiction:

Improveease of dye administrationVSAvoidbackground signals

Core Design Contradiction:

Ease of operationVSObject-generated harmful factors

Solution Approach 1:

The system extracts and removes background signals from hepatic tissue by training a machine learning model to identify and suppress these unwanted signals while preserving the fluorescence signals from structures of interest such as the biliary tree

Inventive Principle:

Principle #2Taking out (Extraction)

Solution Approach 2:

The system changes the parameter of fluorescence signal intensity by applying different suppression levels to different regions of the image, reducing background signal intensity while maintaining or enhancing the intensity of target structure signals

Inventive Principle:

Principle #35Parameter changes

2Object-generated harmful factors

If fluorescent dye is administered 15 hours preoperatively, then background signals are minimized and biliary tree visualization is improved, but patient requires overnight admission which is impractical

Engineering Contradiction:

Improvebackground signalsVSAvoidpatient admission time

Core Design Contradiction:

Object-generated harmful factorsVSLoss of time

Solution Approach 1:

The system converts the harmful background signals from early dye administration into a beneficial situation by using machine learning to identify and suppress these signals, thereby achieving clear visualization without requiring prolonged patient admission

Inventive Principle:

Principle #22Blessing in disguise (Convert harm into benefit)

3Loss of time

If early fluorescence imaging is performed, then the imaging can be done during surgery without delay, but the low contrast between target structures and background reduces measurement precision

Engineering Contradiction:

Improveimaging delayVSAvoidcontrast precision

Core Design Contradiction:

Loss of timeVSMeasurement precision

Solution Approach 1:

The system enhances contrast precision by dynamically adjusting signal intensities across different image regions, suppressing background areas and enhancing target structure areas to achieve high measurement precision in early fluorescence imaging

Inventive Principle:

Principle #35Parameter changes

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 clear visualization of surgical structures by reducing background noise and enhancing the visibility of objects of interest, even when they have lower fluorescence intensity than the background, thereby improving surgical precision and reducing the need for manual threshold adjustments.

Implementation Method 1

identifying a background area in at least one of the visible-light image and the fluorescence medical image by providing at least one of the visible-light image and the fluorescence medical image to one or more trained machine-learning models

Methodology Applied

Scientific EffectMachine learning pattern recognition:

Implementation Method 2

enhancing the fluorescence medical image by suppressing a plurality of pixels in the fluorescence medical image that correspond to the identified background area

Methodology Applied

Scientific EffectSignal suppression:

Implementation Method 3

fluorescence imaging can be used in many surgical procedures to assist in intraoperative decision-making

Methodology Applied

Scientific EffectFluorescence: Fluorescence

Data Source

PatentUS20240354943A1Methods and systems for generating enhanced fluorescence imaging data

Publication Date: 2024.10.24 STRYKER CORP

AI summary

The present disclosure relates generally to medical imaging, and more specifically to machine-learning techniques to generate enhanced fluorescence images of a subject (e.g., to aid a surgery, to aid diagnosis and treatment of diseases). The system can receive a visible-light image of the subject; identify a background area in the visible-light image by providing the visible-light image to one or more trained machine-learning models; and enhance the fluorescence medical image by suppressing a plurality of pixels in the fluorescence medical image that correspond to the identified background area in the visible-light image relative to the rest of the pixels in the fluorescence medical image.