A Hybrid Retrieval Augmented Generation Approach for Explainable Clinical Decision Support Using Vector Databases and Best Matching 25
Abstract
Abstract Clinical decision making in modern healthcare industry faces problems due to the growing amount of medical knowledge, scattered data sources, and the lack of patient centered information. Traditional Clinical Decision Support Systems heavily depend on rules and keyword searches, which always lack an understanding of context and personalization. This paper puts forth a Clinical Decision Support System that is assisted by Retrieval Augmented Generation (RAG) method. It fuses Electronic Health Records with unstructured medical literature. The system uses a hybrid search approach that fuses BM25 lexical search with BioBERT based semantic search. The results are combined using Reciprocal Rank Fusion. Additionally, the system has risk prediction for early disease diagnosis. Experimental evaluation of the retrieval system on roughly 182,000 medical records achieved a Precision@5 of 0.81, a Recall@10 of 0.85, and an MRR of 0.83, surpassing BM25 independent and dense baselines by up to 9.5%. This framework boosts quality of evidence and improves reasoning based on individual patients. It helps make timely decisions and increase reliability in complex, high risk situations.
Read the paper
Related studies
- Knowledge Management Perspective of Generative Artificial Intelligence
- Navigating Generative Artificial Intelligence Promises and Perils for Knowledge and Creative Work
- Characterizing the evolving research on enterprise content management
- Large language models and multimodal foundation models for precision oncology
- From Generative AI to Extended Reality: Multidisciplinary Perspectives on the Challenges, Opportunities, and Future of Educational Computing
- Generative Artificial Intelligence: A New Engine for Advancing Environmental Science and Engineering
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.