The Community Focus Channel on VJHemOnc is an independent medical education platform, supported with funding from Johnson & Johnson (Gold). Supporters have no influence on the production of content. The levels of sponsorship listed are reflective of the amount of funding given.
iwNHL 2025 | AI in practice: gaps in knowledge and decentralized clinical trials
Jason Westin, MD, FACP, The University of Texas MD Anderson Cancer Center, Houston, TX, comments on the gaps in knowledge when using artificial intelligence (AI) in clinical practice. He emphasizes the need for oncologists to gain expertise in AI tools and hopes to see them integrated into medical record systems to streamline clinical trials. This interview took place at the 22nd International Workshop on Non-Hodgkin Lymphoma (iwNHL 2025), held in Cambridge, MA.
These works are owned by Magdalen Medical Publishing (MMP) and are protected by copyright laws and treaties around the world. All rights are reserved.
Transcript
I think there’s a big knowledge gap in AI in general for the public who does not work on AI. Obviously, there are people who’ve dedicated their careers to being at the forefront of developing novel AI models, and AI models are beginning to start training the next AI model, and so the human expertise in that field is there, but many, most oncologists are not well-versed in that. They’ve kind of played with ChatGPT or they use it to help summarize a paper, but they don’t really have full knowledge of all of the tools that it could use in clinical trials as well as in drug development...
I think there’s a big knowledge gap in AI in general for the public who does not work on AI. Obviously, there are people who’ve dedicated their careers to being at the forefront of developing novel AI models, and AI models are beginning to start training the next AI model, and so the human expertise in that field is there, but many, most oncologists are not well-versed in that. They’ve kind of played with ChatGPT or they use it to help summarize a paper, but they don’t really have full knowledge of all of the tools that it could use in clinical trials as well as in drug development. And I would specifically hope that we start to see this being used in what we call decentralized clinical trials, meaning that you could open a study theoretically at hundreds of sites. And instead of having a research nurse or a data coordinator or the infrastructure that we typically think is required for clinical trials to meet each of those sites, you could have an AI tool embedded in a medical record system that effectively acts as the clinical trial staff and reminds the physician to order the lab on this date, that collects the data, that calls the patient to remind them to do various things. I think that the field is going to need to evolve quite rapidly. And so physicians will need to learn about how these tools can be not just a chatbot, but something that actually influences how we practice medicine.
This transcript is AI-generated. While we strive for accuracy, please verify this copy with the video.
Read more...
