Style Neutralization Model for Medical Image Consistency
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Medical images acquired from different devices and techniques vary in imaging characteristics, leading to inconsistent performance when input into artificial intelligence-based diagnosis support programs, affecting diagnosis accuracy.
Innovation Solution
A deep learning-based method and apparatus for medical image style neutralization, utilizing a style neutralization deep learning model that includes inverse-transformation and imitation deep learning models to standardize imaging characteristics, converting input images into neutralized images suitable for AI diagnosis support programs by adjusting parameters such as tube voltage, detection quantum efficiency, and post-processing methods.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If medical images from different devices and techniques are directly input to AI diagnosis support programs, then device versatility and imaging technique diversity are maintained, but diagnosis support performance varies and consistency deteriorates
Solution Approach 1:
The patent introduces a style transfer model as an intermediary between medical imaging devices and AI diagnosis support programs. This model receives images from various devices with different imaging characteristics and transforms them into a standardized format, enabling consistent AI diagnosis performance across multiple device types and imaging techniques while preserving the original device versatility
Solution Approach 2:
The style transfer model changes imaging parameters such as contrast, brightness, and texture characteristics to standardize the appearance of medical images from different devices. By adjusting these visual parameters, the system maintains adaptability to various imaging devices while ensuring consistent diagnosis support performance across all inputs
2Reliability
If style transfer models are trained with large datasets from multiple devices, then model generalization improves, but training data requirements and system complexity increase
Solution Approach 1:
The patent segments the style transfer system into multiple specialized models, each trained on data from a specific device type or imaging technique. This segmentation allows each model to be trained on smaller, device-specific datasets while collectively providing broad generalization capability, reducing the complexity of training a single large-scale model
Data Source
AI summary
Disclosed is a method of deep learning-based medical image style neutralization for generating a neutralized image to be input to an artificial intelligence-based diagnosis support program includes: obtaining a medical image for processing from an outside; and generating a neutralized image by inputting the medical image for the processing to a style neutralization deep learning model trained in advance to neutralize imaging characteristics of the medical image for the processing, wherein the style neutralization deep learning model includes a plurality of style neutralization deep learning models, and the style neutralization deep learning model corresponding to the imaging characteristics of the medical image for the processing performs the neutralization of the medical image for the processing.