CT Image Prompt Reliability Screening for Medical AI Answers

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Generative models in the medical field often produce unreliable and incoherent answers, leading to potential harm to users due to hallucinations.

Innovation Solution

An X-ray CT apparatus and medical information processing system that includes a generative model, an X-ray tube, and processing circuitry to generate and process CT image data, extract named entities, and determine answer reliability by comparing vectors and thresholds, ensuring coherence with user context.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Extent of automation

If generative models are used to provide medical information, then the automation and efficiency of medical information processing is improved, but the reliability and accuracy of the information deteriorates due to hallucinations

Engineering Contradiction:

Improveautomation of medical information processingVSAvoidreliability of medical information

Core Design Contradiction:

Extent of automationVSReliability

Solution Approach 1:

The system implements a feedback mechanism where the reliability determination unit continuously monitors and evaluates the generative model's output. It extracts named entities from both the prompt and answer, compares them against medical knowledge bases, and provides feedback signals to assess reliability. This feedback loop enables the system to detect hallucinations and maintain reliability while preserving automation.

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent introduces an intermediary reliability determination unit that acts as a mediator between the generative model and the final medical information output. This intermediary component extracts named entities, compares them with reference information from medical databases, and determines reliability before the information is delivered to users, thus resolving the contradiction between automation and reliability.

Inventive Principle:

Principle #24Intermediary (Mediator)

2Productivity

If generative models are used to generate medical answers, then the productivity of medical information processing is improved, but the accuracy of the information deteriorates

Engineering Contradiction:

Improveproductivity of medical information processingVSAvoidaccuracy of medical information

Core Design Contradiction:

ProductivityVSMeasurement precision

Solution Approach 1:

The system performs preliminary actions by extracting named entities from the prompt and answer before final accuracy determination. It pre-processes the information by identifying key medical terms, symptoms, and diagnoses, then uses these extracted entities to compare against medical knowledge bases. This preliminary action enables efficient accuracy checking without compromising productivity.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent replaces manual verification mechanisms with automated computational methods. Instead of requiring human experts to verify each medical answer, the system uses computational entity extraction, vector comparisons, and automated reliability determination to assess accuracy, thereby maintaining high productivity while ensuring precision through machine-based verification.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

3Reliability

If reliability determination is performed by comparing named entities and vectors, then the accuracy of medical information is improved, but the device complexity increases

Engineering Contradiction:

Improveaccuracy of medical informationVSAvoidcomplexity of processing system

Core Design Contradiction:

ReliabilityVSDevice complexity

Solution Approach 1:

The system segments the reliability determination process into distinct functional modules: named entity extraction unit, vector comparison unit, and reliability determination unit. Each module handles a specific aspect of the verification process independently. This segmentation reduces overall system complexity by allowing each component to be optimized and maintained separately while achieving high accuracy through coordinated operation.

Inventive Principle:

Principle #1Segmentation

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

Improves the accuracy and reliability of generative model answers in the medical field by assessing similarity and context, reducing health risks associated with unreliable information.

Implementation Method 1

an X-ray tube (51) that generates an X-ray

Methodology Applied

Scientific EffectX-ray generation: X-Ray

Implementation Method 2

an X-ray detector (52) that detects an X-ray emitted from the X-ray tube (51) and having passed a subject (P)

Methodology Applied

Scientific EffectX-ray detection: Photoelectric Effect

Data Source

PatentUS20250359830A1X-ray computed tomography apparatus, medical information processing apparatus, and medical information processing method

Publication Date: 2025.11.27 CANON KK

AI summary

An X-ray computed tomography (CT) apparatus according to an embodiment includes an X-ray tube, an X-ray detector, and processing circuitry. The X-ray tube generates an X-ray. The X-ray detector detects the X-ray emitted from the X-ray tube and having passed a subject. The processing circuitry reconstructs CT image data based on an output from the X-ray detector. The processing circuitry generates a prompt to be input to a generative model, the prompt including observation information about the CT image data. The processing circuitry obtains an answer including medical information about the subject from the generative model, in response to an input of the prompt to the generative model. The processing circuitry outputs reliability of the answer based on the prompt and the answer.