AI Engine Cognifying Unstructured Patient Data via Knowledge Graph
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current healthcare systems face inefficiencies in processing and analyzing unstructured patient data, leading to time-consuming reviews of electronic medical records (EMRs) and potential misdiagnoses due to the overwhelming amount of information, which wastes computing and network resources.
Innovation Solution
A cognitive intelligence platform that uses artificial intelligence to cognify unstructured data by identifying indicia such as phrases, predicates, and keywords, comparing them to a knowledge graph representing health-related information, and generating cognified data through structural similarity, thereby summarizing patient conditions and providing actionable insights.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If physicians review all unstructured EMR data manually, then diagnostic thoroughness is improved, but time consumption and resource waste increase
Solution Approach 1:
The system extracts only the most relevant information from unstructured EMR data by identifying indicia (phrases, predicates, keywords, cardinals, numbers, concepts) and comparing them against a knowledge graph. This extraction process filters out unnecessary information while preserving diagnostically valuable data, allowing physicians to focus on condensed, actionable insights rather than reviewing entire EMR documents.
Solution Approach 2:
An artificial intelligence engine acts as an intermediary between unstructured EMR data and physicians. The engine processes raw data through multiple stages: identifying indicia, comparing with knowledge graphs, detecting structural similarities, and generating cognified data. This intermediary transforms overwhelming unstructured information into organized, clinically relevant summaries that maintain diagnostic accuracy while reducing review time.
2Loss of information
If all unstructured patient data is processed and stored, then information completeness is improved, but computing resource consumption increases
Solution Approach 1:
The system extracts only essential information elements from unstructured data by identifying specific indicia such as medical phrases, predicates, keywords, numerical values, and conceptual terms. This selective extraction maintains information completeness for diagnostically relevant data while eliminating processing of redundant or irrelevant content, thereby reducing computing resource requirements.
Solution Approach 2:
The system applies different processing qualities to different portions of data based on their clinical relevance. High-priority indicia that affect diagnosis receive thorough processing and validation against the knowledge graph, while less critical information receives minimal processing. This local quality approach ensures information completeness where needed while conserving computing resources elsewhere.
3Measurement precision
If unstructured data is analyzed in detail, then diagnostic accuracy is improved, but operational efficiency decreases
Solution Approach 1:
The system performs preliminary analysis by pre-processing unstructured EMR data to identify indicia and compare them against the knowledge graph before physician review. This preliminary action generates pre-processed cognified data that highlights potential diagnoses, anomalies, and clinically significant findings, allowing physicians to make accurate diagnoses more quickly without performing detailed analysis of all raw data.
Solution Approach 2:
The system creates a simplified copy of the essential information from unstructured EMR data in an organized, structured format. Instead of requiring physicians to analyze the original complex unstructured data, the system generates a condensed representation that preserves diagnostically critical information while eliminating formatting complexity, thereby improving operational efficiency without sacrificing diagnostic accuracy.
Data Source
AI summary
A method includes receiving, at an artificial intelligence engine, a corpus of data for a patient, where the corpus of data includes a set of strings of characters. The method also includes identifying, in the set of strings of characters, indicia including a phrase, a predicate, a keyword, a subject, an object, a cardinal, a number, a concept, or some combination thereof. The method also includes comparing the indicia to a knowledge graph representing known health related information to generate a possible health related information pertaining to the patient. The method also includes identifying, using a logical structure, a structural similarity of the possible health related information and a known predicate in the logical structure. The method also includes generating, by the artificial intelligence engine, cognified data based on the structural similarity.