Increasing the Efficiency of the Use of Patient Information Leaflets by Using Retrieval Augmented Generation

Abstract

This paper introduces a Retrieval-Augmented Generation (RAG) system specifically designed for enhancing the accessibility and comprehension of medical information from patient information leaflets documents. Leveraging state-of-the-art technologies such as Optical Character Recognition (OCR), vector embeddings, hybrid search mechanisms combining semantic and full-text search, and Large Language Models (LLMs) like GPT-3.5 turbo, the system efficiently processes and responds to natural language queries. By integrating these components into a cohesive architecture, the RAG system facilitates accurate retrieval of medical data and generates responses that are not only precise but also formatted to be easily understood by laypersons. The effectiveness of the RAG system was evaluated through a series of real-world case studies, which demonstrated its ability to provide reliable, contextually relevant medical advice, thereby significantly improving users' access to essential health information. Insights gained from these studies indicate critical areas for future enhancement, particularly in user interaction and system feedback integration. This work underscores the potential of advanced AI tools to transform information accessibility in healthcare, making critical medical information more approachable for the public.

Keywords

• Retrieval Augmented Generation (RAG)
• AI in Medicine
• Medical Technology
• Large Language Models (LLM)
• OpenAI
• GPT-3.5 turbo

References

1. [1] Tian, S., Jin, Q., Yeganova, L., Lai, P. T., Zhu, Q., Chen, X., ... & Lu, Z. (2024). Opportunities and challenges for ChatGPT and large language models in biomedicine and health. Briefings in Bioinformatics, 25(1), bbad493.
2. [2] Jin, Q., Leaman, R., & Lu, Z. (2023). Retrieve, summarize, and verify: how will ChatGPT affect information seeking from the medical literature?. Journal of the American Society of Nephrology, 34(8), 1302-1304.
3. [3] Lewis, P., Perez, E., Piktus, A., Petroni, F., Karpukhin, V., Goyal, N., ... & Kiela, D. (2020). Retrieval-augmented generation for knowledge-intensive nlp tasks. Advances in Neural Information Processing Systems, 33, 9459-9474.
4. [4] Zhao, P., Zhang, H., Yu, Q., Wang, Z., Geng, Y., Fu, F., ... & Cui, B. (2024). Retrieval-augmented generation for ai-generated content: A survey. arXiv preprint arXiv:2402.19473.
5. [5] Alsentzer, E., Murphy, J. R., Boag, W., Weng, W. H., Jin, D., Naumann, T., & McDermott, M. (2019). Publicly available clinical BERT embeddings. arXiv preprint arXiv:1904.03323.
6. [6] Mann, B., Ryder, N., Subbiah, M., Kaplan, J., Dhariwal, P., Neelakantan, A., ... & Amodei, D. (2020). Language models are few-shot learners. arXiv preprint arXiv:2005.14165, 1.
7. [7] Borgeaud, S., Mensch, A., Hoffmann, J., Cai, T., Rutherford, E., Millican, K., ... & Sifre, L. (2022, June). Improving language models by retrieving from trillions of tokens. In International conference on machine learning (pp. 2206-2240). PMLR.
8. [8] Zhu, X., Lin, T., Anand, V., Calderwood, M., Clausen-Brown, E., Lueck, G., ... & Wu, C. (2023, April). Explicit and Implicit Semantic Ranking Framework. In Companion Proceedings of the ACM Web Conference 2023 (pp. 326-330).

9. [9] Starker, E. (2023). Azure Cognitive Search: Outperforming vector search with hybrid retrieval and ranking capabilities. Microsoft. https://techcommunity.microsoft.com/t5/ai-azure-ai-services/azure-cognitive-search-outperforming-vector-search-with-hybrid/m-p/3931019.
10. [10] Andersson, H. (2024). RETRIEVAL-AUGMENTEDGENERATION WITH AZURE OPEN AI.
11. [11] Bruch, S., Gai, S., & Ingber, A. (2023). An analysis of fusion functions for hybrid retrieval. ACM Transactions on Information Systems, 42(1), 1-35.
12. [12] Chen, X., Gao, C., Chen, C., Zhang, G., & Liu, Y. (2024). An Empirical Study on Challenges for OpenAI Developers. arXiv preprint arXiv:2408.05002.
13. [13] Frisoni, G., Mizutani, M., Moro, G., & Valgimigli, L. (2022, December). Bioreader: a retrieval-enhanced text-to-text transformer for biomedical literature. In Proceedings of the 2022 conference on empirical methods in natural language processing (pp. 5770-5793).