The volume of medical research has grown significantly, making it challenging for researchers to stay up-to-date on recent findings. For instance, the number of cancer-related publications listed in the National Library of Medicine’s PubMed citation database has more than doubled since 2005. To help researchers sift through this vast amount of data, some federal agencies are employing large language model (LLM) functionality.
The National Cancer Institute (NCI) has developed NanCI – Connecting Scientists, a stand-alone mobile and web application launched in 2024. NanCI summarizes and highlights key research findings and recommends relevant papers and researchers to connect with, based on areas of interest identified by biomedical and cancer researchers, says Nastaran Zahir, acting director of the Center for Cancer Training and branch director of the Cancer Training Branch at NCI.
LLMs in Medical Research
NanCI is one of the most visible applications of large language models, according to Zahir. The application provides a valuable model for how these technologies can be applied in a practical and user-centered way. Medical research-associated LLM use has expanded in recent years due to the models’ increasingly complex architecture and the availability of healthcare and biomedical data.
Wes Anderson, quantitative medicine scientist at the Critical Path Institute, notes that the current generation of LLMs can make connections between different words in a sentence and be trained to respond to prompts and follow instructions. These models are being used to address bottlenecks that involve language and knowledge synthesis, Anderson says.
Read Also: Doctors praise AI for restoring joy in medicine
After NanCI‘s introduction, NCI added Ask NanCI, a feature that lets researchers interact with scientific content in a conversational way. Zahir explains that researchers can ask questions, such as “Tell me the major outcome of this paper,” and Ask NanCI will use large language modeling to generate a response.
Google Cloud provides the underlying cloud infrastructure and access to LLMs that power the Ask NanCI function, Zahir says. The NanCI application is available as an iOS app and web application that’s accessible on any device, including Android smartphones.
The U.S. Department of Energy-sponsored Argonne National Laboratory has also launched Argo, a custom interface that provides secure access to large language models such as Google’s Gemini and OpenAI’s GPT-5.4, says Matthew Dearing, the lab’s AI for operations technical lead.
The laboratory’s goal is to provide the core technology that researchers need to leverage generative AI and LLMs for data analysis, content generation, and other objectives, Dearing says. Communication between LLMs, users, and any software they build occurs on the lab’s network, ensuring secure access to sensitive data.
Security was also a consideration for the Food and Drug Administration when building its LLM-based Elsa app. Jeremy Walsh, Chief AI Officer, notes that the agency had to centralize its internal divisions’ separate data silos, which included consolidating 40 application and submission systems into one platform.
Read Also: How basic care boosts patient access before chatbots
The Elsa app, whose infrastructure involves Amazon Web Services and Google Cloud components, allows researchers to analyze applications and identify potential compliance violations. Currently, 80% of FDA employees use Elsa to complete work such as reviewing thousands of direct-to-consumer ad submissions.
Wes Anderson notes that LLMs can potentially produce problematic responses and other inaccuracies. Some agencies, including the FDA, have issued guidance on how to approach the different risk levels that the models’ use can present.
As the technology advances, understanding where failures can exist, and how to ensure models are continuously evaluated to confirm their research output is accurate, can help build credibility, Anderson says. The use of LLMs in medical research is becoming increasingly prevalent, with applications such as NanCI and Elsa demonstrating their potential to expedite research processes and improve productivity.
Implementing LLMs in Medical Research
Amazon Web Services and Google Cloud components are used in the infrastructure of the Elsa app, and the app can be accessed with an internet connection and an agency-issued Dell laptop.