AI-Steered Endoscope for Precise Upper GI Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Endoscopic procedures for capturing images of the upper gastrointestinal tract are highly dependent on manual manipulation by medical specialists, leading to difficulties in obtaining precise and clear images due to limitations in adjusting the position and angle of the endoscope, especially in areas with many curves, which affects the efficiency and accuracy of diagnosis.
Innovation Solution
An endoscopic device controlled by an artificial intelligence (AI) neural network that acquires images, detects body parts, calculates relative position and pose information, and generates control signals to steer the endoscope's front-end portion to specific body parts, using pre-trained models and torque feedback to adjust the endoscope's position and angle.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual manipulation by medical specialist is used to adjust endoscope position and angle, then the endoscope can be controlled to capture images, but the precision and clarity of images deteriorate due to limitations in adjusting position and angle especially in curved areas
Solution Approach 1:
The patent replaces manual mechanical manipulation with an automated image processing system that uses artificial intelligence to detect body parts and calculate relative position information, thereby substituting the mechanical control system with an intelligent vision-based system to achieve precise image acquisition without manual intervention limitations
Solution Approach 2:
The endoscopic device performs self-positioning and self-adjustment by automatically detecting body parts in real-time images and calculating the necessary position corrections, enabling the system to service itself without external manual manipulation, thus overcoming the precision limitations of manual operation
2Productivity
If manual manipulation is used to adjust endoscope position and angle, then the procedure can be performed, but the efficiency and accuracy of diagnosis deteriorate due to difficulty in precise adjusting
Solution Approach 1:
The system continuously captures real-time images, detects body parts, calculates relative position information, and generates control signals to adjust the endoscope position, creating a closed-loop feedback system that automatically corrects positioning errors and maintains optimal viewing angles, thereby improving both efficiency and accuracy of diagnosis
Solution Approach 2:
The system pre-calculates the necessary position adjustments by detecting body parts and computing relative position information before actual image capture, allowing the endoscope to be proactively positioned at optimal locations rather than reactively adjusted during the procedure, thus enhancing diagnostic efficiency and accuracy
3Measurement precision
If AI neural network control is implemented to detect body parts and generate control signals, then positioning accuracy is improved, but device complexity increases
Solution Approach 1:
The controller integrates multiple functions including image acquisition, AI-based body part detection, relative position calculation, and endoscope control into a single multi-functional device, reducing the need for separate specialized components and thereby managing complexity while maintaining high positioning accuracy through integrated intelligent control
Data Source
AI summary
Provided are a medical image device and an automated control technique, particularly, an endoscopic device for capturing an image of an upper gastrointestinal tract using an endoscope and a method of controlling the endoscopic device. The method of controlling an endoscopic device includes acquiring an image of an upper gastrointestinal tract from an image sensor, detecting at least one first body part from the image, based on a pre-trained model, calculating relative position information between the detected at least one first body part and a front-end portion of the endoscopic device, generating, based on the relative position information, a first control signal for steering the front-end portion to correspond to the at least one first body part, and transmitting the first control signal to a driver.