Leveraging Artificial Intelligence, Natural Language Processing, and Natural Language Generation in Medical Writing
DOI:
https://doi.org/10.55752/amwa.2023.180Abstract
Medical writing is a process that generates a variety of documents in the biomedical domain, including but not limited to clinical reports, regulatory reports, protocol documents, patient narratives, plain language summaries, and so on. Medical writing is complex and time-consuming because a writer must refer to multiple sources, sift through a large volume of documents, maintain data integrity, perform review of literature, do interpretation of results, summarize, and so on.
These challenges can be addressed and minimized substantially by adopting artificial intelligence, specifically cognitive search, natural language processing (NLP), and natural language generation (NLG) models and other techniques. Given the recent advances in language models for NLG, the time is ripe for a product in the medical writing domain that integrates and automates search capabilities, provides cognitive processing, and generates content using NLG.
This white paper takes scientific manuscript writing as an example to provide insights into the way NLP and NLG can augment, automate, and expedite the process of writing a wide variety of biomedical documents. It looks at the current limitations of technology and ways to address those. Finally, it provides recommendations on how these technologies can be used to create a single system or product. Such an approach has the potential to expand into multiple areas in the biomedical domain, with medical writing as the first challenge.
Downloads
Published
How to Cite
Issue
Section
License
Copyright Notice
Authors of manuscripts accepted for publication retain copyright in their work but are required to grant to AMWA exclusive worldwide first publication rights and further grant a nonexclusive license for other uses of the manuscript for the duration of their copyright as outlined in the Agreement on Authors' Rights and Responsibilities. In the same agreement, authors are required to confirm that the material submitted for publication, both in its paper and electronic versions, including reproductions of other works (eg, photographs, maps, etc.) does not infringe upon any existing copyright.
Any reproduction within two years following the date of publication requires the permission of the Editor-in-Chief of the AMWA Journal.
Contact managingeditor@amwa.org for all requests for reuse.