Clinical Explainable AI Visualization for Trusting Levels and Risk Indexes

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Medical personnel lack the ability to interpret and verify the accuracy of predicting results generated by artificial intelligence methods in clinical medicine, leading to potential incorrect medical decisions and treatments.

Innovation Solution

An explainable artificial intelligence system and method that utilizes a processor to read parameter datasets, machine learning models, and model explainable programs to generate visualization information, including trusting levels and risk indexes, thereby enhancing the interpretability and reliability of predicting results.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If artificial intelligence method or machine learning method is used to assist medical personnel in making treatment decisions, then productivity and accuracy of medical diagnosis is improved, but the black box characteristic prevents medical personnel from interpreting or verifying the predicting result

Engineering Contradiction:

Improveaccuracy of medical diagnosisVSAvoidinterpretability of predicting result

Core Design Contradiction:

Measurement precisionVSLoss of information

Solution Approach 1:

The patent introduces an explainable artificial intelligence module as an intermediary between the machine learning model and medical personnel. This module generates visualization information including trusting levels and risk indexes that mediate the communication between the black box model and human users, making the model's reasoning process interpretable without modifying the underlying machine learning algorithms

Inventive Principle:

Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent uses visual representation with different colors to indicate trusting levels of predicting results. The visualization information displays color-coded indicators that represent the reliability and confidence of AI-generated predictions, allowing medical personnel to quickly assess the trustworthiness of each prediction at a glance

Inventive Principle:

Principle #32Color changes

2Productivity

If AI generating predicting result without explanation is used, then productivity is improved, but reliability of medical decision-making deteriorates due to inability to verify accuracy

Engineering Contradiction:

Improveefficiency of treatment decisionVSAvoidreliability of medical decision

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The patent implements a feedback mechanism where the explainable AI module continuously provides visualization information including trusting levels and risk indexes back to the medical personnel. This feedback loop allows clinicians to verify AI predictions against clinical judgment and provide corrections, thereby improving the reliability of medical decisions while maintaining high productivity

Inventive Principle:

Principle #23Feedback

Solution Approach 2:

The patent performs preliminary analysis by generating visualization information and trusting levels before the medical personnel makes final treatment decisions. The system pre-calculates risk indexes and provides confidence assessments in advance, allowing clinicians to review and verify predictions before committing to treatment plans

Inventive Principle:

Principle #10Preliminary action

3Device complexity

If machine learning model provides only predicting result without explanation, then device complexity is reduced, but ease of operation deteriorates as medical personnel cannot interpret the results

Engineering Contradiction:

Improvesimplicity of AI systemVSAvoidease of interpreting predicting result

Core Design Contradiction:

Device complexityVSEase of operation

Solution Approach 1:

The patent segments the complex AI system into two distinct components: a machine learning model that generates predictions and an explainable AI module that generates visualization information. This segmentation allows the system to remain simple in terms of model architecture while providing comprehensive explanation capabilities through the separate visualization layer

Inventive Principle:

Principle #1Segmentation

Solution Approach 2:

The explainable AI module serves as an intermediary that translates complex model outputs into easily interpretable visualization information. This intermediary layer converts difficult-to-interpret model predictions into intuitive visual representations including trusting levels and risk indexes, greatly improving ease of operation without adding complexity to the core machine learning model

Inventive Principle:

Principle #24Intermediary (Mediator)

Data Source

PatentUS20250378959A1Explainable artificial intelligence method applied to clinical medicine and system thereof and non-transitory computer readable recording medium

Publication Date: 2025.12.11 CHINA MEDICAL UNIVERSITY(TW)

AI summary

An explainable artificial intelligence method applied to clinical medicine includes reading a parameter dataset, a machine learning model, a model explainable program and a plurality of clinical index range values, the parameter dataset includes a plurality of parameters; inputting the parameter dataset into the machine learning model to generate a predicting result; executing the model explainable program to the machine learning model, to calculate a plurality of important values and a plurality of risk indexes; determining whether one of the parameters being out of one of the clinical index range values; comparing the parameters and the risk indexes to generate a risk information; comparing the parameters and the clinical index range values, and dividing the parameters into a plurality of trusting levels; and integrating the parameters, the important values, the risk information and the trusting levels into a visualization information.