Diagnosis Support System Using Conversion Function for Missing Examination Values
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Specialists in diagnosis and treatment departments face difficulties in early detection of diseases outside their areas of expertise, and existing support technologies struggle to adapt across different departments due to variations in examination types.
Innovation Solution
A diagnosis and treatment support system that includes a storage apparatus with a trained model and processing circuitry capable of calculating a conversion function to derive missing examination values, generating support information, and inferring patient states using machine learning techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If specialists use existing support technology for early detection of diseases outside their expertise, then early detection capability is improved, but the system cannot be easily adapted across different diagnosis and treatment departments due to variations in examination types
Solution Approach 1:
The system creates a universal support framework that can be applied across different diagnosis and treatment departments. The machine learning model is trained on examination data from multiple departments and diseases, enabling it to provide early detection support for various disease types beyond a single specialist's expertise area while maintaining department-specific characteristics through the shared model architecture
2Measurement precision
If a support system is designed for a specific diagnosis and treatment department, then it achieves high accuracy for that department, but it cannot be used easily for other departments with different examination types
Solution Approach 1:
The system segments the support functionality into a shared machine learning model component and department-specific application components. The core detection algorithm is segmented as a universal model trained on multi-department data, while allowing for department-specific adaptations, enabling both high accuracy for specific departments and usability across other departments
Solution Approach 2:
The system changes parameters by training the machine learning model on diverse examination data from multiple departments with different examination types. The model learns to adapt to varying examination parameters and characteristics, maintaining accuracy across different departmental contexts through parameter adaptation rather than requiring separate models for each department
Data Source
AI summary
A diagnosis and treatment support system according to an embodiment has a storage apparatus and processing circuitry. The storage apparatus stores a trained model that infers information related to a state of a patient from an examination value for a predetermined examination item. Based on correlations between examination values for plural examination items, the processing circuitry calculates a conversion function that enables statistical derivation of a possible examination value for another examination item from an examination value or values for one or plural examination items, and generates, based on an inference obtained by inputting an examination value for an examination item included in diagnosis and treatment information on a target patient and the examination value for the predetermined examination item derived by the conversion function from the examination value into the trained model, diagnosis and treatment support information for the target patient.