AI Endoscopy Guidance for Complete Gastroscopy Coverage
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing gastroscopy procedures are time-consuming and prone to incomplete coverage of examination sites and missed suspicious areas due to the lack of efficient guidance for digestive endoscopists, leading to reduced quality and potential risks for patients.
Innovation Solution
An auxiliary endoscopy system that includes an anatomic-site identification module to map observation points, an attention degree recording module to determine observation sequences, and a guide module to direct the endoscope to target stations, utilizing an AI-driven neural network for precise site identification and route planning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If digestive endoscopists perform gastroscopy manually without assistance, then they can complete the examination, but it takes a long time and is prone to incomplete coverage of examination sites and missed suspicious areas
Solution Approach 1:
The system enables self-service by having the endoscopy device automatically identify anatomical sites, determine observation sequences, and generate guidance routes without requiring manual intervention from the endoscopist for these functions. The endoscope itself performs the analysis and guidance generation, reducing dependency on operator experience.
Solution Approach 2:
The system implements feedback by continuously monitoring the endoscopy process and providing real-time guidance based on identified anatomical sites and observation sequences. The guidance information is dynamically adjusted based on the actual endoscopy progress, ensuring complete coverage of examination sites.
2Reliability
If digestive endoscopists increase their workload to ensure complete examination coverage, then examination quality improves, but the endoscopists become overloaded and quality deteriorates
Solution Approach 1:
The system transfers the complexity of site identification, sequence determination, and route planning from the endoscopist to the automated system. The endoscope device handles these complex tasks independently, allowing endoscopists to focus on actual examination procedures without mental overload.
Solution Approach 2:
The system acts as an intermediary between the endoscopist and the examination process. It mediates by providing structured guidance information that simplifies the endoscopist's decision-making process, reducing cognitive load while maintaining examination quality.
3Ease of operation
If less experienced digestive endoscopists perform gastroscopy, then they can conduct examinations, but examination sites are often missed or suspicious areas fail to be detected
Solution Approach 1:
The system enables less experienced endoscopists to perform examinations at a higher level by having the endoscope device autonomously identify anatomical sites and generate guidance routes. This compensates for the operator's limited experience in site recognition and examination sequencing.
Solution Approach 2:
The system provides real-time feedback to less experienced endoscopists through automated guidance information that highlights important anatomical sites and suggested observation sequences. This feedback mechanism compensates for their lack of experience by continuously directing their attention to critical areas.
Data Source
AI summary
The present application relates to an endoscope auxiliary examination system, a method, an apparatus, and a storage medium. The system comprises: an anatomical part identification module, used to perform frame-by-frame identification on collected endoscope examination images, determine every observation point part corresponding to each frame of the endoscope examination images, and map each observation point part to a preset site part; an attention recording module, used to determine a site part observation sequence according to a recorded observation time of each observation point part and an observation completeness of each site part determined by the anatomical part identification module; and a guidance module, used to determine a target site to be moved to, according to a current site part corresponding to a current frame endoscope examination image determined by the anatomical part identification module, the site part observation sequence and a preset physical depth of the current site part, and guiding endoscopic examination according to the current site part and the target site. According to the present application, an endoscopist is guided to operate an endoscope according to the most efficient route.