Research on Auxiliary Decision-Making Combining 
Medical Knowledge Graphs and RAG

Yihang Zhang

doi:10.70711/aitr.v2i8.6626

Research on Auxiliary Decision-Making Combining Medical Knowledge Graphs and RAG

Yihang Zhang

Abstract

The quality of medical decision-making is critical for patient care, especially given the pressure on physicians to process vast information quickly and the challenges posed by uneven resource distribution and inexperienced doctors. This paper presents a knowledge graphbased system that integrates Retrieval-Augmented Generation (RAG) to provide diagnostic references, enhance empathy, collect feedback,
and optimize treatment plans. By using extractive summarization and a real-time reward function, the system reduces retrieval time and ensures high-quality inputs for improved outcomes.

Keywords

Medical auxiliary decision-making; Deep learning; RAG model; Knowledge graph; NLP

Full Text:

PDF

Included Database

References

[1] Wang, Y., et al. (2021). "KG-GPT: Enhancing Generative Pre-trained Models with Knowledge Graphs." Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing (EMNLP).

[2] Liu, H., et al. (2022). "Joint Training of Graph Neural Networks and Generative Models for Context-Aware Dialogue Systems." Transactions on Knowledge and Data Engineering (TKDE).

[3] Zhang, L., et al. (2023). "Multimodal Knowledge Graph Enhanced Generation for Vision-Language Tasks." Proceedings of the 2023

International Conference on Learning Representations (ICLR).

[4] Zhang, Y., et al. (2021). "Medical Entity Recognition Using BERT-Based Models for Enhanced Information Retrieval." Journal of Biomedical Informatics.

[5] Li, X., et al. (2023). "Entity-Aware Retrieval Augmentation in Healthcare: Combining NER with Knowledge Graphs." Proceedings of

the 2023 Conference on Health Informatics and AI Applications.

[6] Fangyuan Xu, et al. "RECOMP: Improving Retrieval-Augmented LMs with Compression and Selective Augmentation." arXiv preprint,

2023, arXiv:2310.04408v1 (accessed: 2023-10-15).

DOI: http://dx.doi.org/10.70711/aitr.v2i8.6626

Refbacks

There are currently no refbacks.