Medical Image Acquisition Information from Image and Non-Image Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for automating medical image processing for diagnosis are inadequate due to insufficient and inaccurate documentation of image acquisition parameters, leading to inefficient resource consumption and user workload, especially when dealing with diverse image and non-image data from various modalities and acquisition technologies.
Innovation Solution
A computer-implemented method using two trained functions, one for image data and one for non-image data, to determine image acquisition information, combining their outputs to provide a consolidated and accurate image acquisition information for medical images, leveraging neural networks and natural language processing to extract relevant parameters.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If image acquisition information is extracted only from documented protocols, then the process is simple, but the accuracy and completeness of image acquisition information is insufficient
Solution Approach 1:
The patent segments the image acquisition information extraction into two independent components: (1) extracting information from documented protocols via text processing, and (2) inferring information from image data via trained functions. These segmented approaches are then combined to produce comprehensive and accurate image acquisition information, resolving the contradiction between extraction simplicity and information accuracy.
Solution Approach 2:
The patent introduces trained functions as intermediary components that bridge the gap between raw image data and meaningful image acquisition information. These trained functions act as mediators that analyze image characteristics and infer acquisition parameters, thereby improving accuracy without requiring direct access to complete documentation.
2Productivity
If automated processing is performed with incomplete image acquisition information, then the workflow is fast, but the processing quality and diagnostic accuracy deteriorate
Solution Approach 1:
The patent performs preliminary extraction of image acquisition information from both protocols and image data before the main processing workflow begins. By pre-processing and consolidating information from multiple sources in advance, the system ensures that complete and accurate acquisition information is available when automated processing starts, thereby maintaining both speed and quality.
3Loss of information
If multiple data sources are integrated to improve information accuracy, then the completeness of image acquisition information is enhanced, but the system complexity increases
Solution Approach 1:
The patent merges information from multiple independent sources (protocol documentation and image data analysis) into a unified image acquisition information set. By combining these complementary information sources, the system achieves comprehensive information coverage while managing complexity through modular integration of extraction and inference components.
Data Source
Figure 1~2
Figure 3~4
Figure 5
AI summary
Computer-implemented methods and systems for providing an image acquisition information of a medical image are provided. The methods and systems implement a plurality of steps. One step is directed to obtaining the medical image. Another step is directed to extracting image data from the medical image. Another step is directed to obtaining non-image data associated with the image data of the medical image. Another step is directed to providing a first trained function configured to determine an image acquisition information based on non-image data of medical images. Another step is directed to providing a second trained function configured to determine an image acquisition information based on image data associated with image data of the medical image. Another step is directed to determining a first image acquisition information by applying the first trained function to the non-image data. Another step is directed to determining a second image acquisition information by applying the second trained function to the image data. Another step is directed to determining the image acquisition information based on the first image acquisition information and the second image acquisition information. Another step is directed to providing the image acquisition information.