Weighted Graph Relevance Scoring for Patient Information Retrieval
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The abundance of patient information in electronic medical records can lead to relevant information being hidden, causing caregivers to spend precious time searching through extensive records to locate vital data.
Innovation Solution
A system utilizing a weighted graph to determine the relevance of historical patient information to a current condition by representing the current condition and historical information as nodes, allowing for the calculation of a relevance score and prioritization of pertinent information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If the full patient medical history is made available to caregivers, then the completeness of information is improved, but the time required to review and locate relevant information increases
Solution Approach 1:
The patent segments the complete patient medical history into discrete, manageable units represented as nodes in a graph structure. Each node contains specific medical information elements (diagnoses, procedures, medications, etc.) that can be independently evaluated for relevance. This segmentation allows the system to process and present information in organized chunks rather than overwhelming full records, resolving the contradiction by maintaining information completeness while enabling efficient selective review.
Solution Approach 2:
The patent introduces a computational intermediary (the graph-based relevance scoring system) that acts as a mediator between the complete medical records and the caregiver. This intermediary automatically analyzes the full patient history, calculates relevance scores based on the current condition, and presents only the most pertinent information. The intermediary preserves all original information while filtering out irrelevant content, thus maintaining completeness without requiring the caregiver to review everything manually.
2Loss of information
If all patient records are presented equally, then information completeness is maintained, but the ability to quickly identify critical information deteriorates
Solution Approach 1:
The patent applies local quality by assigning different relevance scores to different portions of the patient medical history based on their specific relationship to the current condition. Instead of treating all information uniformly, the system evaluates each node (medical record element) individually and assigns a quality metric (relevance score) that reflects its local importance. This allows critical information to be prominently displayed while less relevant information is de-emphasized, maintaining data availability while enabling rapid identification of key elements.
Solution Approach 2:
The patent changes the parameter of information presentation by transforming static medical records into a dynamic graph structure with calculable relevance scores. The system modifies how information is parameterized - from uniform record formats to nodes with associated relevance metrics that can be sorted and filtered. This parameter change enables the system to maintain complete information while providing a ranked presentation that accelerates retrieval of critical data based on calculated relevance to the current condition.
Data Source
AI summary
Media, methods, and systems facilitate identification and presentation of medical information that is relevant to a current condition of a patient. The relative prominence of the information is associated with the relative relevance of the information with the current condition. A weighted graph is used to determine the relevance.