Unified Medical Image Analysis via Navigation Trajectory Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing medical image analysis technologies require multiple models for different diseases, leading to low efficiency, high computational cost, and low accuracy, as they can only analyze whole images and not specific disease areas, making them unsuitable for clinical applications.
Innovation Solution
A medical image analysis method that determines a navigation trajectory based on analysis requirements, extracts image blocks along this trajectory using a first learning network, and uses a second learning network to analyze these features, allowing for accurate and efficient analysis of multiple diseases within a unified framework.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If multiple models are used to analyze different diseases, then the analysis coverage is improved, but the device complexity and labor cost increase
Solution Approach 1:
The patent applies a single unified model that can analyze multiple different diseases (breast cancer, lung nodules, bone fractures, etc.) instead of requiring separate specialized models for each disease type. This universal model achieves multi-functionality by processing various medical images and disease types through one system, thereby reducing device complexity while maintaining comprehensive analysis coverage.
Solution Approach 2:
The patent segments the medical image into multiple image blocks along a navigation trajectory, allowing the single model to focus on specific regions of interest for different disease types. This segmentation enables the unified model to efficiently handle multiple disease analyses by processing relevant image blocks rather than entire images, reducing computational complexity while maintaining versatility.
2Reliability
If the whole medical image is analyzed, then the analysis completeness is improved, but the computational cost and time increase
Solution Approach 1:
The patent extracts only the relevant image blocks along the navigation trajectory that contain potential disease information, rather than analyzing the entire medical image. This extraction principle allows the system to maintain analysis completeness for disease detection while significantly reducing computational cost and processing time by focusing only on pertinent regions.
Solution Approach 2:
The patent applies partial action by analyzing only specific image blocks along the navigation trajectory rather than the complete image. This selective analysis of partial regions maintains sufficient diagnostic information for disease detection while improving analysis efficiency and reducing computational resources required.
3Measurement precision
If manual adjustment of model settings for different diseases is performed, then the analysis accuracy is improved, but the labor cost and time increase
Solution Approach 1:
The unified model performs self-adjustment through automatic navigation trajectory generation and image block extraction based on the input medical image and disease type. The system automatically determines which image blocks to analyze and processes them through the single model without requiring manual configuration changes, thereby maintaining high analysis accuracy while eliminating time-consuming manual adjustment operations.
Data Source
AI summary
The present disclosure relates to a medical image analysis method, a medical image analysis device, and a computer-readable storage medium. The medical image analysis method includes receiving a medical image acquired by a medical imaging device; determining a navigation trajectory by performing navigation processing on the medical image based on an analysis requirement, the analysis requirement indicating a disease to be analyzed; extracting an image block set along the navigation trajectory; extracting image features using a first learning network based on the image block set; and determining an analysis result using a second learning network based on the image features and the navigation trajectory.