Medical Image Data Retrieval via Segmentation Indicators
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for retrieving data from large 3D medical image datasets are time-consuming and inefficient, requiring users to navigate through numerous slices to locate areas of interest, especially with increasing spatial resolution leading to larger datasets and longer analysis times.
Innovation Solution
Implementing a method that uses automatic segmentation to identify and represent objects within the dataset through indicators, allowing users to select and retrieve only the relevant data slices associated with these indicators, reducing the need to load the entire dataset.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If the complete data set is retrieved for inspection, then all objects of interest can be accessed, but the retrieval time and data transfer requirements increase significantly
Solution Approach 1:
The patent segments the complete medical image data set into multiple groups based on anatomical regions or diagnostic categories. Each segment can be independently retrieved and displayed, allowing users to access only relevant portions rather than the entire dataset, thus reducing retrieval time while maintaining access to all necessary diagnostic information.
Solution Approach 2:
The patent extracts and displays representative thumbnail images from different segments of the data set simultaneously in an overview display. This allows users to identify and select only the specific segments containing objects of interest, extracting the necessary information without retrieving the complete data set and thereby reducing data transfer requirements and retrieval time.
2Ease of operation
If thumbnail representations are used to overview the data set, then navigation is facilitated, but the thumbnails are based on slice numbering which is not effective for selecting relevant anatomical parts
Solution Approach 1:
The patent applies local quality by organizing thumbnail representations according to anatomical regions or diagnostic categories rather than uniform slice numbering. Each thumbnail group represents a specific anatomical area with relevant metadata, allowing users to quickly identify and navigate to clinically relevant regions without being overwhelmed by irrelevant slices, thus improving navigation efficiency while preserving anatomical context.
Data Source
AI summary
The present invention discloses a method of retrieving a plurality of data slices from a medical image data set (5), the method comprising the steps of: a) displaying an indicator (10, 20) associated with the plurality of data slices; b) selecting the indicator (10, 20) based on a user input; and c) retrieving the plurality of data slices (1, 2) associated with the indicator when said indicator is selected; wherein the association between the indicator and the plurality of slices is based on segmentation of the medical image data set, the indicator representing an object obtained in the segmentation of the medical image data set, the plurality of data slices comprising the object data. The method of the invention reduces the amount of data transfer because it allows for retrieving only those data slices which comprise relevant data relating to the object of interest.