Synapse
⌘+K
Synapse
PulseExploreJournal ClubResearchersJournals
Instagram
HomeJournal ClubExplore
April 15, 2026

Medical Question Answering System Based on Retrieval-augmented Generation

View Full Paper
Ask AI
Bookmark
Share

Authors

CCChengwei Cao

Discussion

Loading...

Member takes

Overview

Demonstrates improved answer accuracy in medical question answering, suggesting RAG's potential in AI systems.

Key Points

  • The research aims to evaluate the effectiveness of Retrieval-Augmented Generation (RAG) in medical question-answering tasks.
  • Constructed a medical knowledge base from a cancer-related subset of a Kaggle healthcare dataset.
  • Conducted experiments using Qwen-Plus and Qwen-Flash models.
  • Evaluated performance based on answer accuracy, source traceability, refusal capability, and retrieval quantity.
  • RAG significantly improved the semantic consistency of model responses.
  • Outperformed the baseline model on the BERTScore F1 metric.
  • Demonstrated strong performance in refusal rate and attribution accuracy.
  • A retrieval quantity of k=5 yielded the best overall performance.

Cite This Study

Chengwei Cao (2026) studied this question.

synapsesocial.com/papers/69df2c62e4eeef8a2a6b1834https://doi.org/10.1051/itmconf/20268401001/pdf
View Full Paper
Ask AI
Bookmark
Share