NY hospital exec: Multimodal LLM assistants will create a “paradigm shift” in patient care

NewYork-Presbyterian’s Medical Director of AI Operations, Dr. Ashley Beecy, highlighted the potential of large language models’ role in developing virtual doctor assistants. Speaking at the AI Impact Tour in New York, Beecy discussed the hospital system’s use of generative AI for tasks like summarizing patient conversations, aiming to create comprehensive assistants for proactive patient care. While the technical capability exists, streamlining workflow processes within the healthcare system and promoting collaboration on a national level are crucial factors for the integration of this technology.

Looking ahead, Beecy envisions a single assistant handling various tasks seamlessly, from scheduling patient visits to clinical diagnostics. NYP employees have shown enthusiasm for generative AI applications, particularly in reducing administrative burdens for doctors by automating note-taking during patient visits. Transitioning to clinical diagnostics, such as screening for heart disease using electrocardiograms, presents more significant challenges but promises to enhance patient care by providing early detection.

Despite the advancements in AI technology, Beecy emphasized the importance of human oversight to mitigate risks associated with errors or overreliance on AI-generated outputs. NYP adopts a cautious approach, prioritizing alignment between those implementing generative AI tools and end-users to ensure successful integration into existing workflows. Surveying patient groups on the transparency of AI usage reflects NYP’s commitment to ethical considerations and patient trust.