Medical Image ROI Contour Formatting for Lower Data Burden
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current medical imaging systems face challenges in efficiently processing large volumes of data due to the overlaying of analysis results on original images, leading to increased resource burden and inefficient data transmission.
Innovation Solution
A method utilizing deep learning to detect regions of interest in medical images, generating format information that defines the representation of these regions, and combining this information with the original image to reduce data volume and improve processing efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If analysis results are overlaid on original medical images and transmitted to output terminals, then visualization quality is improved, but data volume increases and processing resources are burdened
Solution Approach 1:
The patent extracts only the essential analysis result data (contour information, region of interest coordinates, diagnostic findings) from the complete medical image data, separating it from the original image. This allows transmission of minimal data volume while preserving visualization quality by reconstructing the overlay representation from extracted features rather than transmitting the entire overlaid image.
Solution Approach 2:
The patent segments the medical image data into distinct components: original image data, analysis result data (contour information, region markers, diagnostic text), and metadata. By segmenting the data structure, the system transmits only the necessary analysis components separately from the original image, reducing overall data volume while maintaining the ability to reconstruct high-quality visualizations at the output terminal.
2Loss of information
If analysis results are overlaid on original medical images, then diagnostic information is enhanced, but computing resources are consumed
Solution Approach 1:
The patent performs preliminary extraction and structuring of analysis result data during the image processing stage, organizing contour information, region of interest data, and diagnostic findings into a standardized format. This preliminary action reduces the computational burden during transmission and display stages, as the output terminal only needs to reconstruct the visualization from pre-processed data rather than performing complex overlay operations on large datasets.
3Reliability
If large-scale medical images consisting of hundreds of series are processed, then comprehensive diagnosis is achieved, but data processing burden increases
Solution Approach 1:
The patent extracts only the critical diagnostic information from large-scale multi-series medical images, identifying and isolating region of interest data, contour information, and key diagnostic findings. This extraction process maintains comprehensive diagnostic capability by focusing on clinically relevant data while discarding redundant information, thereby significantly reducing the processing burden for large-scale image datasets.
Solution Approach 2:
The patent segments large-scale medical images into multiple series and further divides each series into region of interest segments. By organizing the data hierarchy into image series → regions of interest → contour elements, the system enables selective processing and transmission of only necessary segments, improving productivity when handling comprehensive multi-series diagnostic data.
Data Source
AI summary
According to an exemplary embodiment of the present disclosure, a medical image processing method performed by a computing device is disclosed. The medical image processing method includes: detecting a region of interest in a medical image by using a pre-trained deep learning model; determining contour information for the region of interest; and generating, based on the contour information, format information defining elements that determine representation of the medical image.