Downloads

https://doi.org/10.62836/jcmea.v4i1.040108
Zhaoyan Zhang, Yu Qiao, & Peimin Lu. (2024). Self-Reflective Retrieval-Augmented Framework for Reliable Pharmacological Recommendations. Journal of Computational Methods in Engineering Applications, 4(1), 1–12. https://doi.org/10.62836/jcmea.v4i1.040108
Volume 4, Issue 1, 2024

Self-Reflective Retrieval-Augmented Framework for Reliable Pharmacological Recommendations

Pharmacological recommendations are critical for ensuring patient safety and treatment efficacy, yet traditional methods often struggle with inaccuracies and limited adaptability to new knowledge. To address these challenges, this paper proposes a novel Self-Reflective Retrieval-Augmented Framework for reliable pharmacological recommendations. The framework incorporates three key innovations: a self-reflective mechanism for dynamic error detection and correction, a pharmacological memory bank for long-term reasoning and knowledge accumulation, and a RAG-enhanced retrieval module to dynamically integrate up-to-date external knowledge during recommendation generation. Experiments on datasets from DrugBank and FDA adverse event reporting systems demonstrate that the proposed framework significantly improves recommendation accuracy, with the full model achieving a 92.3% accuracy and outperforming state-of-the-art methods across multiple evaluation metrics. This research provides a robust and adaptive solution for pharmacological recommendation tasks, paving the way for safer and more effective decision-making in healthcare.

Retrieval-Augmented Generation; artificial intelligence; pharmacological recommendations; deep learning

References

  1. Yao S, Yu Z, Du N, et al. React: Synergizing Reasoning and Acting in Language Models. arXiv 2023; arXiv:2210.03629.
  2. Shinn N, Labash J, Savinov N. Reflexion: Language Agents with Verbal Reinforcement Learning. arXiv 2023; arXiv:2303.11366.
  3. He S, Nie Y, Chen Z, et al. Meddr: Diagnosis-Guided Bootstrapping for Large-Scale Medical Vision-Language Learning. arXiv 2024; arXiv:2404.15127.
  4. Kumar Y, Marttinen P. Improving Medical Multi-Modal Contrastive Learning with Expert Annotations. arXiv 2024; arXiv:2403.10153.
  5. Sun L, Zhao J, Han M, et al. Fact-Aware Multimodal Retrieval Augmentation for Accurate Medical Radiology Report Generation. arXiv 2024; arXiv:2407.15268.
  6. Tao Y, Ma L, Yu J, et al. Memory-Based Cross-Modal Semantic Alignment Network for Radiology Report Generation. IEEE Journal of Biomedical and Health Informatics 2024; 28: 4145–4156.
  7. Yuan Z, Jin Q, Tan C, et al. Ramm: Retrieval-Augmented Biomedical Visual Question Answering with Multi-Modal Pre-Training. In Proceedings of the 31st ACM International Conference on Multimedia, Ottawa, ON, Canada, 29 October–3 November 2023; pp. 547–556.
  8. Chen Z, Xu C, Qi Y, et al. Mllm Is a Strong Reranker: Advancing Multimodal Retrieval-Augmented Generation via Knowledge-Enhanced Reranking and Noise-Injected Training. arXiv 2024; arXiv:2407.21439.
  9. Gao Y, Xiong Y, Gao X, et al. Retrieval-Augmented Generation for Large Language Models: A Survey. arXiv 2023;arXiv:2312.10997.
  10. Qu X, Chen Q, Wei W, et al. Alleviating Hallucination in Large Vision-Language Models with Active Retrieval Augmentation. arXiv 2024; arXiv:2408.00555.
  11. Wu Y, Zhu J, Xu S, et al. Ragtruth: A Hallucination Corpus for Developing Trustworthy Retrieval-Augmented Language Models. arXiv 2023; arXiv:2401.00396.
  12. Wei J, Wang X, Schuurmans D, et al. Chain of thought Prompting Elicits Reasoning in Large Language Models. arXiv 2022; arXiv:2201.11903.
  13. Yao S, Zhao J, Yu Z, et al. Tree of thoughts: Deliberate Problem Solving with Large Language Models. arXiv 2023; arXiv:2305.10601.
  14. Xia P, Zhu K, Li H, et al. Rule: Reliable Multimodal Rag for Factuality in Medical Vision Language Models. arXiv 2024; arXiv:2407.05131.
  15. Xia P, Zhu K, Li H, et al. Mmed-rag: Versatile Multimodal Rag System for Medical Vision Language Models. arXiv 2024; arXiv:2410.13085.
  16. Li C, Wong C, Zhang S, et al. Llava-Med: Training a Large Language-and-Vision Assistant for Biomedicine in One Day. Advances in Neural Information Processing Systems 2023; 36: 28541–28564.
  17. Moor M, Huang Q, Wu S, et al. Med-Flamingo: A Multimodal Medical Few-Shot Learner. PMLR 2023; 225: 353–367.
  18. Schrouff J, Harris N, Koyejo S, et al. Diagnosing failures of fairness transfer across distribution shift in real-world medical settings. Advances in Neural Information Processing Systems 2022; 35: 19304–19318.
  19. Chen Z, Du Y, Wen Z, et al. MJ-Bench: Is Your Multimodal Reward Model Really a Good Judge for Text-to-Image Generation? arXiv 2024; arXiv:2407.04842.
  20. Liu H, Li C, Li Y, et al. Improved Baselines with Visual Instruction Tuning. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, Seattle, WA, USA, 16–22 June 2024; pp. 26296–26306.
  21. Liu H, Li C, Wu Q, et al. Visual Instruction Tuning. Advances in Neural Information Processing Systems 2023; 36: 34892–34916.
  22. Xia P, Yu X, Hu M, et al. Hgclip: Exploring Vision-Language Models with Graph Representations for Hierarchical Understanding. arXiv 2023; arXiv:2311.14064.
  23. Zhou Y, Cui C, Yoon J, et al. Analyzing and Mitigating Object Hallucination in Large Vision-Language Models. arXiv 2023; arXiv:2310.00754.
  24. Jiang X, Zhang R, Xu Y, et al. Think and Retrieval: A Hypothesis Knowledge Graph Enhanced Medical Large Language Models. arXiv 2023; arXiv:2312.15883.
  25. Lozano A, Fleming SL, Chiang CC, et al. Clinfo. ai: An Open-Source Retrieval-Augmented Large Language Model System for Answering Medical Questions Using Scientific Literature. In Pacific Symposium on Biocomputing, 2024; World Scientific: Singapore, 2023; pp. 8–23.
  26. Royer C, Menze B, Sekuboyina A. Multimedeval: A benchmark and a toolkit for evaluating medical vision-language models. arXiv 2024; arXiv:2402.09262.
  27. Xia P, Chen Z, Tian J, et al. Cares: A Comprehensive Benchmark of Trustworthiness in Medical Vision Language Models. arXiv 2024; arXiv:2406.06007.

Supporting Agencies

  1. Funding: This research has been partly funded by the National Natural Science Foundation of China (NSFC) through awards 61872364 and 61972221. Any opinions, findings, and conclusions or recommendations expressed in this paper are those of the authors and do not necessarily reflect the views of the NSFC.