Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Unlocking Electronic Health Records: A Hybrid Graph RAG Approach to Safe Clinical AI for Patient QA

Samuel Thio, Matthew Lewis, Spiros Denaxas, Richard JB Dobson

TL;DR

The paper addresses the challenge of safe, context-grounded clinical AI for patient-level QA on EMR data. It proposes MediGRAF, a hybrid Graph RAG system that merges Neo4j-based structured querying (Text2Cypher) with vector embeddings for unstructured narratives. Using 10 patients from MIMIC-IV, it reports 5,973 nodes and 5,963 relationships, with 100% recall for factual queries and a mean expert quality of 4.25/5 for complex inferences, with zero safety violations. The work demonstrates that hybrid graph-grounding improves retrieval safety and informativeness and lays groundwork for NHS-scale deployment via CogStack.

Abstract

Electronic health record (EHR) systems present clinicians with vast repositories of clinical information, creating a significant cognitive burden where critical details are easily overlooked. While Large Language Models (LLMs) offer transformative potential for data processing, they face significant limitations in clinical settings, particularly regarding context grounding and hallucinations. Current solutions typically isolate retrieval methods focusing either on structured data (SQL/Cypher) or unstructured semantic search but fail to integrate both simultaneously. This work presents MediGRAF (Medical Graph Retrieval Augmented Framework), a novel hybrid Graph RAG system that bridges this gap. By uniquely combining Neo4j Text2Cypher capabilities for structured relationship traversal with vector embeddings for unstructured narrative retrieval, MediGRAF enables natural language querying of the complete patient journey. Using 10 patients from the MIMIC-IV dataset (generating 5,973 nodes and 5,963 relationships), we generated enough nodes and data for patient level question answering (QA), and we evaluated this architecture across varying query complexities. The system demonstrated 100\% recall for factual queries which means all relevant information was retrieved and in the output, while complex inference tasks achieved a mean expert quality score of 4.25/5 with zero safety violations. These results demonstrate that hybrid graph-grounding significantly advances clinical information retrieval, offering a safer, more comprehensive alternative to standard LLM deployments.

Unlocking Electronic Health Records: A Hybrid Graph RAG Approach to Safe Clinical AI for Patient QA

TL;DR

The paper addresses the challenge of safe, context-grounded clinical AI for patient-level QA on EMR data. It proposes MediGRAF, a hybrid Graph RAG system that merges Neo4j-based structured querying (Text2Cypher) with vector embeddings for unstructured narratives. Using 10 patients from MIMIC-IV, it reports 5,973 nodes and 5,963 relationships, with 100% recall for factual queries and a mean expert quality of 4.25/5 for complex inferences, with zero safety violations. The work demonstrates that hybrid graph-grounding improves retrieval safety and informativeness and lays groundwork for NHS-scale deployment via CogStack.

Abstract

Electronic health record (EHR) systems present clinicians with vast repositories of clinical information, creating a significant cognitive burden where critical details are easily overlooked. While Large Language Models (LLMs) offer transformative potential for data processing, they face significant limitations in clinical settings, particularly regarding context grounding and hallucinations. Current solutions typically isolate retrieval methods focusing either on structured data (SQL/Cypher) or unstructured semantic search but fail to integrate both simultaneously. This work presents MediGRAF (Medical Graph Retrieval Augmented Framework), a novel hybrid Graph RAG system that bridges this gap. By uniquely combining Neo4j Text2Cypher capabilities for structured relationship traversal with vector embeddings for unstructured narrative retrieval, MediGRAF enables natural language querying of the complete patient journey. Using 10 patients from the MIMIC-IV dataset (generating 5,973 nodes and 5,963 relationships), we generated enough nodes and data for patient level question answering (QA), and we evaluated this architecture across varying query complexities. The system demonstrated 100\% recall for factual queries which means all relevant information was retrieved and in the output, while complex inference tasks achieved a mean expert quality score of 4.25/5 with zero safety violations. These results demonstrate that hybrid graph-grounding significantly advances clinical information retrieval, offering a safer, more comprehensive alternative to standard LLM deployments.
Paper Structure (37 sections, 2 equations, 5 figures, 2 tables)

This paper contains 37 sections, 2 equations, 5 figures, 2 tables.

Table of Contents

  1. Introduction
  2. Significance and Clinical Integration
  3. Related Work
  4. Graph Databases in Healthcare Data Management
  5. Large Language Models and the Context Gap
  6. Retrieval-Augmented Generation (RAG) and GraphRAG
  7. Bridging the Interaction Gap: Text2Cypher
  8. The Case for Hybrid Retrieval Architectures
  9. Methodology
  10. Data Source and Preprocessing
  11. Graph Database Schema Design
  12. Vector Embedding Implementation
  13. MediGRAF Pipeline Architecture
  14. Query Complexity Classification
  15. Evaluation Framework and Metrics
  16. ...and 22 more sections

Figures (5)

  • Figure 1: Neo4j schema of MIMIC-IV data with clinical nodes and relationships
  • Figure 2: A workflow diagram showing the MediGRAF architecture from query input through graph retrieval to LLM augmentation and response generation
  • Figure 3: Overview of the MediGRAF system interface showing the main query input area and configurable retrieval parameters. The left sidebar provides filtering options for patient selection, note type filtering, and advanced settings for both vector search (document retrieval limits) and graph database queries (maximum records to return).
  • Figure 4: Comparison of mean expert evaluation scores across four dimensions. Error bars represent standard deviation. Note the significant divergence in 'Relevance & Conciseness' scores.
  • Figure 5: Proposed deployment pipeline illustrating the integration of MediGRAF with the NHS CogStack ecosystem. CogStack handles real-time ingestion and NLP extraction (via MedCAT) to dynamically populate the Neo4j graph, enabling MediGRAF to query live clinical data.