Iterative AI Model for Infarction Region Detection in CT
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for diagnosing cerebral infarction and large vessel occlusion in non-contrast CT images are inefficient, leading to delayed treatment due to the difficulty in accurately specifying infarction regions and occlusion parts, which can result in poor prognosis.
Innovation Solution
An information processing apparatus and method that uses a processor to iteratively update and derive infarction regions and occlusion parts in non-contrast CT images using trained discriminative models, incorporating anatomical and clinical information, and symmetrical brain region data to enhance accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If non-contrast CT image is used for initial diagnosis, then diagnosis speed is improved, but measurement precision of infarction region and occlusion part deteriorates
Solution Approach 1:
The patent segments the diagnosis task into multiple stages: initial rapid screening using non-contrast CT, followed by targeted acquisition of additional images (contrast CT or MRI) only when infarction or LVO is suspected. This segmentation allows fast initial diagnosis while ensuring accurate confirmation only when necessary.
Solution Approach 2:
The system performs preliminary analysis of the non-contrast CT image using AI algorithms to predict infarction regions and LVO parts before final diagnosis. This preliminary action identifies high-risk cases that require further imaging, enabling early intervention planning while maintaining diagnostic accuracy.
2Measurement precision
If contrast CT or MRI is acquired after non-contrast CT, then measurement precision is improved, but loss of time increases
Solution Approach 1:
The patent implements a dynamic diagnostic pathway that adapts to each patient's specific case. The system continuously updates the diagnosis as new information becomes available, adjusting the need for additional imaging based on the initial AI analysis results and clinical presentation, thereby minimizing unnecessary delays.
Solution Approach 2:
The system uses feedback from the AI analysis of non-contrast CT images to determine whether additional imaging is necessary. The AI provides predictive information about infarction and LVO that feeds back into the clinical decision-making process, allowing physicians to make informed decisions about whether to proceed with contrast CT or MRI, thus reducing unnecessary imaging and time loss.
3Productivity
If AI-based automatic extraction is used, then productivity is improved, but measurement precision of occlusion part deteriorates due to difficulty in distinguishing from similar structures
Solution Approach 1:
The patent introduces an intermediary verification step where the AI's preliminary identification of occlusion parts is cross-checked against additional imaging data (contrast CT or MRI) when available. This intermediary process helps distinguish true occlusion parts from similar structures like calcification, improving precision while maintaining the productivity benefits of AI automation.
Solution Approach 2:
The system merges multiple data sources including non-contrast CT, contrast CT, and MRI images to comprehensively identify occlusion parts. By combining information from different imaging modalities, the system overcomes the limitations of single-modality AI analysis and achieves both high productivity and precision in occlusion part specification.
Data Source
AI summary
A processor acquires a medical image and a first disease region in the medical image, derives a second disease region related to the first disease region in the medical image based on the medical image and the first disease region, updates the first disease region based on the medical image and the second disease region, updates the second disease region based on the medical image and the updated first disease region, and repeats update of the first disease region and update of the second disease region until a predetermined end condition is satisfied.