Endoscopic Tile Image Assessment for Consistent Pathology Review
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The assessment of pathological conditions based on endoscopic images is subjective and varies among specialists, and biopsy procedures carry risks such as bleeding and infectious diseases.
Innovation Solution
A medical support system using a derivation device that cuts affected site images into tile shapes and derives assessment values through machine learning, presented on a display device to assist physicians in making more accurate diagnoses.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If visual judgment by specialists is used to assess pathological conditions, then diagnostic capability is obtained, but assessment variation occurs depending on individual specialists
Solution Approach 1:
The patent replaces the mechanical/visual judgment system of specialists with an automated image processing system using machine learning. The derivation device automatically derives assessment values from endoscopic images, substituting human visual assessment with algorithm-based analysis to eliminate inter-observer variation while maintaining diagnostic capability
Solution Approach 2:
The patent introduces an intermediary derivation device that acts as a mediator between the endoscopic image and the final diagnosis. This device processes images through machine learning algorithms to generate objective assessment values, serving as a bridge that translates visual information into quantifiable metrics without direct human interpretation
2Measurement precision
If biopsy is performed for medical examination, then definitive diagnosis is obtained, but risk of bleeding and infectious diseases occurs due to perforation
Solution Approach 1:
The patent creates a virtual copy of the biopsy process through image analysis. Instead of physically extracting tissue samples, the system derives assessment values from endoscopic images using machine learning, producing diagnostic information that replicates the value of histological examination without physical intrusion
Solution Approach 2:
The patent extracts diagnostic information directly from endoscopic images without extracting physical tissue samples. The derivation device extracts relevant features and patterns from images to generate assessment values, eliminating the need for biopsy while preserving diagnostic capability
3Reliability
If machine learning determiner is used to derive assessment values from images, then assessment consistency is improved, but device complexity increases
Solution Approach 1:
The patent segments the complex machine learning system into distinct functional units: an image acquisition unit, a derivation device with cut-out unit and assessment derivation unit, and a display device. This modular segmentation manages complexity by separating concerns while maintaining the overall machine learning functionality
Data Source
AI summary
There is provided a medical support system including: a derivation device that derives an assessment value for an affected site based on an affected site image obtained by imaging the affected site; and a display device that presents the assessment value to a user, in which the derivation device includes: a cut-out unit that cuts out the affected site image as a plurality of tile images having tile shapes; and an assessment derivation unit that derives a tile assessment value representing an assessment of the affected site in the plurality of the tile images by using a determiner obtained by machine learning.