CT Image Generation Control Using Learning-Based Diagnostic Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The generation of unnecessary images in CT systems leads to increased storage and network traffic, and the on-demand image generation process is inefficient, making timely interpretation difficult.
Innovation Solution
An information processing apparatus that uses a learning device to analyze CT images and control the generation of images based on the analysis results, suppressing unnecessary image creation and enabling rapid display of relevant images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If images of a plurality of image types are automatically generated and stored in PACS, then diagnosis efficiency is improved by providing multiple image types in a short period, but storage capacity is excessively consumed and network traffic increases
Solution Approach 1:
The system performs preliminary analysis using a learning device to predict which image types will be useful for diagnosis before generating them. This preliminary action filters out unnecessary image types, so only predicted-useful images are generated and stored in PACS, reducing storage consumption while maintaining diagnosis efficiency
Solution Approach 2:
Instead of generating all possible image types (excessive action), the system generates only a subset of image types that are predicted to be useful (partial action). This selective generation approach reduces the quantity of stored image data while still providing sufficient diagnostic information
2Quantity of substance
If on-demand image generation is implemented, then storage and network resources are conserved by generating only requested images, but interpretation efficiency deteriorates due to time delays in image generation
Solution Approach 1:
The system performs preliminary generation of image types predicted to be useful for diagnosis, before the user actually requests them. This preliminary action stores pre-generated images in PACS, so when users request images, they are already available immediately, eliminating interpretation delays while avoiding generation of unnecessary image types
3Reliability
If all generated images are transmitted to PACS via network, then complete image availability is ensured for all diagnostic needs, but network traffic and storage strain increase
Solution Approach 1:
The system performs preliminary analysis using the learning device to predict which image types will be useful, and only transmits these predicted-useful images to PACS via the network. This preliminary filtering reduces network traffic and storage strain while ensuring that all transmitted images are reliably available for actual diagnostic needs
Data Source
Figure 1
Figure 2
Figure 3
AI summary
Provided are an information processing apparatus, an information processing method, and a program that can suppress the generation of unnecessary images in a CT apparatus and provide images useful for diagnosis in a time-efficient manner. An information processing apparatus (10) includes a processor (102), in which the processor (102) is configured to: acquire a first image captured by a CT apparatus (30) ; execute an analysis process of the first image using a learning device; cause the CT apparatus (30) to execute a generation process of a second image corresponding to a result of the analysis process; and display at least one of the first image, the result of the analysis process, or the second image on a display device (124).