Endoscope Image Recognition for Previously Visited Location Navigation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing endoscopic navigation systems rely on external devices and medical personnel's experience, making it difficult to accurately identify locations within human cavities, increasing the risk of improper examination.

Innovation Solution

A method and image processing device that builds a database of previously observed locations using endoscope images, employing machine learning to compare and recognize these locations, reducing dimensionality to enhance accuracy and provide navigation assistance.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If external devices such as echo devices are used to assist navigation, then navigation accuracy is improved, but device complexity increases

Engineering Contradiction:
Improvenavigation accuracyVSAvoiddevice complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent combines the navigation assistance function directly into the endoscope system by integrating an image processing device that automatically compares current images with stored reference images. This merging eliminates the need for separate external echo devices while maintaining navigation accuracy through automated image recognition and comparison algorithms.

Inventive Principle:
Principle #5Merging (Combining)

Solution Approach 2:

The patent replaces mechanical/external navigation assistance devices with an automated image processing system that uses computer vision algorithms. The image processing device automatically captures, stores, and compares images without requiring external echo devices or manual operator intervention, substituting mechanical navigation aids with intelligent software-based navigation.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Device complexity

If medical personnel rely on experience to navigate the endoscope, then device complexity is reduced, but location identification accuracy deteriorates

Engineering Contradiction:
Improvedevice complexityVSAvoidlocation identification accuracy
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The patent implements a self-service navigation system where the endoscope system automatically performs location identification by comparing images against a database of reference images. The image processing device autonomously captures images, determines whether locations have been previously visited, and provides navigation guidance without relying on medical personnel's subjective experience or manual tracking.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent incorporates a feedback mechanism where the image processing device continuously compares current endoscope images with stored reference images and provides real-time information about previously visited locations. This feedback loop enables automated location recognition and navigation assistance, replacing reliance on operator experience with objective, data-driven navigation information.

Inventive Principle:
Principle #23Feedback

3Reliability

If the endoscope is moved back to earlier positions for re-examination, then examination thoroughness is improved, but examination time increases

Engineering Contradiction:
Improveexamination thoroughnessVSAvoidexamination time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The patent implements preliminary action by automatically storing reference images at various locations during the examination process. When the endoscope returns to a previously visited position, the image processing device quickly compares the current image with the pre-stored reference image to confirm location recognition, enabling rapid re-examination without time-consuming manual tracking or navigation.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent replaces manual navigation and location tracking with an automated image recognition system. The image processing device automatically identifies previously visited locations by comparing images, eliminating the need for medical personnel to manually track the endoscope's position and remember previously examined areas, thereby reducing examination time while maintaining thoroughness.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentEP4191531B1An endoscope image processing device
Publication Date: 2026.01.28 AMBU AS
  • EP4191531B1 patent drawingFigure 1a~1b
  • EP4191531B1 patent drawingFigure 2
  • EP4191531B1 patent drawingFigure 3~4

AI summary

Disclosed is a method of recognising previously observed locations during an endoscopic procedure. The method comprises obtaining a stream of images from an image capturing device of an endoscope during the endoscopic procedure. The method further comprises, during the endoscopic procedure, building a database of previously observed locations based on images of the stream of images. Additionally, the method comprises continuously comparing images of the stream of images with the database of previously observed locations to determine if an image of the stream of images is recorded from a previously observed location. Further disclosed is an image processing device for recognising previously observed locations during an endoscopic procedure comprising a processing unit configured to perform the method, a display unit connectable to the processing unit, an endoscope system comprising an endoscope and the image processing device, and a computer program product.