Patient-Specific Medical Knowledge Graphs for Disease Classification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing medical ontology and graph structures fail to adequately account for variations in patient background factors, leading to inconsistent medical judgment outcomes for patients with similar diseases.
Innovation Solution
A medical information processing apparatus that utilizes a medical knowledge graph with patient background information to modify and vary patient graphs, incorporating a machine learning model to estimate medical judgment information, such as disease classification, by training parameters based on patient-specific factors.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a unified medical knowledge graph structure is used for all patients, then the system complexity is reduced and ease of operation is improved, but measurement precision of medical judgment deteriorates due to inability to account for individual patient variations
Solution Approach 1:
The medical knowledge graph is segmented into two distinct components: a standardized base graph containing universal medical knowledge applicable to all patients, and patient-specific variation graphs that capture individual differences. This segmentation allows the system to maintain a simple unified structure for common medical knowledge while adding personalized adjustments only where necessary, thereby improving measurement precision without proportionally increasing overall complexity.
Solution Approach 2:
The system applies local quality by maintaining a standardized graph structure for the majority of medical knowledge while introducing patient-specific variations only in localized areas where individual differences matter. The variation graphs modify specific nodes and edges relevant to each patient's background factors, rather than restructuring the entire graph, thus improving accuracy for individual cases without globally increasing complexity.
2Measurement precision
If patient background information is incorporated into the medical knowledge graph, then measurement precision of disease classification is improved, but device complexity increases due to additional data processing requirements
Solution Approach 1:
Patient background information is pre-processed and integrated into variation graphs before being applied to the base medical knowledge graph. The system prepares patient-specific adjustment data in advance, organizing background factors and their corresponding graph modifications beforehand. This preliminary action reduces the complexity of real-time data processing during medical judgment, as the integration work is performed during data preparation rather than during the actual classification process.
3Adaptability or versatility
If a standardized medical ontology is used for all patients, then ease of operation and consistency are improved, but adaptability to individual patient characteristics deteriorates
Solution Approach 1:
The system merges a standardized base medical knowledge graph with patient-specific variation graphs to create a personalized medical judgment system. The base graph provides consistent, standardized medical knowledge applicable to all patients, while the variation graphs are merged in to incorporate individual patient characteristics. This combination approach maintains the simplicity and consistency of standardized ontology while adding the necessary adaptability for patient-specific customization without requiring a complete redesign of the system structure.
Data Source
AI summary
A storage device stores a medical knowledge graph including nodes corresponding to medical care events, and edges indicative of a relationship between the nodes. A graph feature of the medical knowledge graph is expressed by a mathematical model characterized by patient background information. A processing circuitry obtains patient background information relating to a background factor of one or more target patients, computes variation of the graph feature relating to the target patient, based on the patient background information of the target patient and the mathematical model, and displays the variation of the graph feature on a display device.