AI Gastroscopic Image Analysis with Anatomical Mapping and Lesion Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing gastroscopic image analysis methods are limited to diagnosing gastric lesions and cannot accurately identify the anatomical location of the stomach or adjacent regions, leading to potential misdiagnosis and inefficiencies in cancer detection due to reliance on clinician proficiency and image acquisition precision.
Innovation Solution
An artificial intelligence-based gastroscopic image analysis method that utilizes multiple image classification models to automatically recognize and report the anatomical location of the stomach and adjacent regions by analyzing gastroscopic images, including training data for each model and segmenting specific parts of the stomach for precise lesion detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple image classification models are used to recognize anatomical locations, then measurement precision of lesion location is improved, but device complexity increases
Solution Approach 1:
The patent divides the gastroscopic image analysis into multiple specialized image classification models, each trained to recognize specific anatomical locations (esophagus, stomach, duodenum). This segmentation allows each model to focus on particular regions, improving overall measurement precision of lesion location while managing complexity through modular architecture
Solution Approach 2:
The system employs multiple image classification models that can universally identify various anatomical locations across different gastroscopic images. Each model serves multiple functions by recognizing different anatomical structures, allowing the system to handle diverse diagnostic scenarios with a unified multi-model framework
2Productivity
If AI training models are used for automated lesion detection, then productivity of diagnosis is improved, but reliability of diagnosis may worsen due to potential misdiagnosis
Solution Approach 1:
The system performs preliminary classification of anatomical locations using trained image classification models before final lesion diagnosis. This preliminary action of identifying the specific gastrointestinal region provides context for subsequent diagnosis, improving reliability while maintaining high productivity through automated processing
Solution Approach 2:
The patent implements a feedback mechanism where the AI model's classification results are used to guide and refine the diagnosis process. The system continuously learns from training data and adjusts its predictions, providing feedback loops that improve both the speed and accuracy of automated lesion detection over time
Data Source
AI summary
Disclosed is an artificial intelligence (AI)-based gastroscopic image analysis method including a step of configuring training data and training the plurality of image classification models, a step of observing or photographing respective parts of a stomach from the oral cavity and the laryngopharynx to the second part of the duodenum using a gastroscopic probe, a step of automatically classifying and recognizing the anatomical location of the stomach with respect to an image captured during photographing using the plurality of image classification models, automatically storing or reporting the location of a lesion, and a step of segmenting an image of a specific part in the captured gastroscopic image for each region by a region segmentation model and outputting the same number of segmented maps as the number of target classes as the result of segmentation.