Medical Image Processing Using Exam-Selected Learning Data
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In medical information processing apparatuses, the increased amount of learning result data obtained by machine learning leads to prolonged processing times for reading and extracting data, often including unnecessary data, which degrades processing quality and operability, making it difficult to quickly display suitable medical images.
Innovation Solution
An image processing apparatus that selectively obtains learning result data based on examination information, using a first obtaining means to gather examination info, a readout means to retrieve relevant learning result data from storage, and a processing means to process medical images using the retrieved data, thereby reducing unnecessary data processing and improving efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple images acquired by different imaging apparatuses are integrated, then the evaluation value for determining presence/absence of a lesion improves, but the complexity of image processing and calculation increases
Solution Approach 1:
The patent segments the image integration process by creating separate processing paths for different imaging modalities (first imaging apparatus and second imaging apparatus). Each path processes images independently through its own integration unit, and only the final evaluation is combined. This segmentation reduces processing complexity while maintaining detection accuracy.
Solution Approach 2:
The patent introduces an evaluation information generation unit that acts as an intermediary, converting complex multi-source image data into simplified evaluation values. This intermediary layer handles the complexity of integrating multiple imaging modalities while presenting simplified results to the determination unit, thus resolving the contradiction between accurate lesion detection and processing complexity.
2Reliability
If images from multiple imaging apparatuses are integrated to determine lesion presence, then diagnostic reliability improves, but the time required for image processing increases
Solution Approach 1:
The patent performs preliminary integration of images from different imaging apparatuses in separate integration units before final evaluation. By pre-processing and organizing the multi-source images in advance, the system reduces the time required for final lesion determination while maintaining diagnostic reliability through comprehensive image integration.
Solution Approach 2:
The patent divides the time-consuming image integration process into parallel segments for different imaging modalities. Each integration unit processes images from specific apparatuses independently and simultaneously, reducing total processing time while maintaining the reliability benefits of multi-source integration through the evaluation information generation unit.
3Measurement precision
If comprehensive image integration is performed to reduce misdiagnosis, then diagnostic accuracy improves, but the computational resources and processing complexity increase
Solution Approach 1:
The evaluation information generation unit serves as an intermediary that manages the complexity of comprehensive image integration. It receives integrated images from multiple apparatuses, performs the necessary computational analysis, and generates simplified evaluation values that indicate lesion presence. This intermediary structure maintains high diagnostic accuracy while managing computational complexity in a organized manner.
Solution Approach 2:
The patent creates separate integration units that effectively copy the image integration function for different imaging modalities. Each unit processes images from specific apparatuses using standardized procedures, reducing the overall system complexity by reusing the same integration logic across multiple processing paths while still achieving comprehensive integration for high diagnostic accuracy.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
An image processing apparatus comprises a first obtaining unit that obtains examination information, a readout unit that reads out learning result data selected based on the examination information from a storage unit storing at least one piece of learning result data that has been obtained by machine learning in advance, a second obtaining unit that obtains a medical image obtained based on the examination information obtained by the first obtaining unit, and a processing unit that processes the medical image obtained by the second obtaining unit using learning result data read out by the readout unit.