International Journal of World Medicine, 2026, 7(1); doi: 10.38007/IJWM.2026.070104.
Zihan Zhang
University of Manitoba, Winnipeg, Manitoba, Canada
Early-stage work in drugs and natural products, biomaterials, and medical surface materials often has a mixed and uneven literature foundation. The first task in the prototype described here is to organise the evidence, not to automatically discover it. Link the research question to the domain schema, a traceable Evidence Ledger and a staged agent workflow in the system. It can query and modify the information of candidate paper, build a reading queue, save attribute records with timestamps, flag weak claims, retain PDF-access failures as red flags, and keep track of the status of each processing attempt for follow-up investigation. The above is a Feasibility demonstration of the described workflow. It should not be taken to mean that the extraction model has already been validated on a gold-standard set.
AI-assisted research platform; drugs and natural products; biomaterials; medical surface materials; domain schema; Evidence Ledger; multi-agent workflow
Zihan Zhang. Design and Prototype Validation of an AI-Assisted Evidence Workflow Platform for Early-Stage Biomedical and Materials Research. International Journal of World Medicine (2026), Vol. 7, Issue 1: 28-41. https://doi.org/10.38007/IJWM.2026.070104.
[1] Vamathevan J, Clark D, Czodrowski P, et al. Applications of machine learning in drug discovery and development. Nature Reviews Drug Discovery. 2019;18(6):463-477.
[2] Paul D, Sanap G, Shenoy S, et al. Artificial intelligence in drug discovery and development. Drug Discovery Today. 2021;26(1):80-93.
[3] Williams DF. On the mechanisms of biocompatibility. Biomaterials. 2008;29(20):2941-2953.
[4] Place ES, George JH, Williams CK, et al. Synthetic polymer scaffolds for tissue engineering. Chemical Society Reviews. 2009;38(4):1139-1151.
[5] Li W, Thian ES, Wang M, et al. Surface design for antibacterial materials: From fundamentals to advanced strategies. Advanced Science. 2021;8(19):2100368.
[6] Arciola CR, Campoccia D, Montanaro L. Implant infections: Adhesion, biofilm formation and immune evasion. Nature Reviews Microbiology. 2018;16(7):397-409.
[7] Lewis P, Perez E, Piktus A, et al. Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks. In: Advances in Neural Information Processing Systems. 2020;33:9459-9474.
[8] Wilkinson MD, Dumontier M, Aalbersberg IJJ, et al. The FAIR Guiding Principles for scientific data management and stewardship. Scientific Data. 2016;3:160018.
[9] Robertson S, Zaragoza H. The Probabilistic Relevance Framework: BM25 and Beyond. Foundations and Trends in Information Retrieval. 2009;3(4):333-389.
[10] Johnson J, Douze M, Jegou H. Billion-scale similarity search with GPUs. IEEE Transactions on Big Data. 2019;7(3):535-547.
[11] Devlin J, Chang MW, Lee K, et al. BERT: Pre-training of deep bidirectional transformers for language understanding. In: Proceedings of NAACL-HLT. 2019:4171-4186.
[12] Beltagy I, Lo K, Cohan A. SciBERT: A pretrained language model for scientific text. In: Proceedings of EMNLP-IJCNLP. 2019:3615-3620.
[13] Vaswani A, Shazeer N, Parmar N, et al. Attention is all you need. In: Advances in Neural Information Processing Systems. 2017;30:5998-6008.
[14] Thirunavukarasu AJ, Ting DSJ, Elangovan K, et al. Large language models in medicine. Nature Medicine. 2023;29:1930-1940.
[15] Wei J, Wang X, Schuurmans D, et al. Chain-of-thought prompting elicits reasoning in large language models. In: Advances in Neural Information Processing Systems. 2022;35:24824-24837.
[16] Yao S, Zhao J, Yu D, et al. ReAct: Synergizing reasoning and acting in language models. In: The Eleventh International Conference on Learning Representations (ICLR). 2023.
[17] Elicit. Systematic literature reviews: AI for evidence synthesis. 2026 [cited 2026-06-15]. Available from: https://elicit.com/solutions/systematic-review.
[18] Consensus. Consensus: AI for research. 2026 [cited 2026-06-15]. Available from: https://consensus.app/.
[19] Nicholson JM, Mordaunt M, Lopez P, et al. scite: A smart citation index that displays the context of citations and classifies their intent using deep learning. Quantitative Science Studies. 2021;2(3):882-898.
[20] SciSpace. AI-powered data extraction from research PDFs. 2026 [cited 2026-06-15]. Available from: https://scispace.com/extract-data.