Diagnostic Detection Method Using Inverse Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing diagnostic methods struggle to accurately distinguish between uncertain and non-corresponding information in diagnostic results, making it difficult to identify undefined attributes and properly classify input classes.
Innovation Solution
A detection method and apparatus that execute inference processing using medical care guidelines, determining correspondence between input classes and subject classes, and their negation classes, to output appropriate coping methods, and repeat inference processing for lower-level classes to identify uncertain attributes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If existing diagnostic methods are used to determine diagnostic results, then diagnostic processing can be performed, but it is difficult to accurately distinguish between uncertain and non-corresponding information, leading to reduced measurement precision
Solution Approach 1:
The patent segments the diagnostic determination process into three distinct processing steps: (1) determining correspondence between input class and subject class, (2) determining correspondence with negation classes, and (3) outputting results based on combined determination. This segmentation allows the system to distinguish between uncertain attributes and non-corresponding conditions, thereby improving diagnostic accuracy without creating an unduly complex overall system.
Solution Approach 2:
The patent introduces negation classes that contradict the subject class conditions and determines correspondence with these inverted conditions. By checking both the subject class and its negation, the system can identify whether attributes are uncertain (non-corresponding to subject but also non-corresponding to negation) versus truly non-corresponding, thus improving measurement precision through inverse verification.
2Measurement precision
If inference processing is executed to identify uncertain attributes, then diagnostic accuracy is improved, but processing time increases
Solution Approach 1:
The patent performs preliminary determination of correspondence between input classes and subject classes before final output. By pre-establishing the correspondence relationships and uncertainty status of attributes during the processing steps, the system avoids redundant computations and enables efficient final determination, thus improving diagnostic accuracy while controlling processing time.
Solution Approach 2:
The patent executes inference processing selectively - performing full determination processing only when needed to distinguish uncertain from non-corresponding attributes. The system determines correspondence for subject classes and negation classes, and only performs additional inference processing when attributes are identified as uncertain, avoiding unnecessary processing for clearly corresponding or non-corresponding cases, thus balancing accuracy with processing efficiency.
Data Source
AI summary
A detection method includes: executing first processing that includes determining whether or not at least any of one or more attribute values included in an input class corresponds to any of one or more conditions defined in a subject class; executing second processing that includes determining whether or not at least any of the one or more attribute values included in the input class corresponds to a negation class including one or more second conditions that contradict the one or more conditions defined in the subject class; and executing third processing that includes outputting information relating to the input class determined as non-corresponding by the second processing when both a determination result in the first processing and a determination result in the second processing are non-corresponding.