Exploring artificial intelligence for home hemodialysis knowledge access

Home hemodialysis (HHD) is a complex therapy, and support information is often scattered across manuals, training guides, handbooks, and videos. To explore new ways of improving access to this information, researchers are developing and evaluating an artificial intelligence-driven system designed to bring Fresenius Medical Care (FME)-approved source materials together in a single, searchable resource.

In this Research in Brief, Radomir Popovic, Senior Data Scientist on the Digital Services & Solutions team, and Sundar Ramaswamy, PhD, Director of Product Management Connected Health on the Home Therapy team, discuss the development of this experimental technology and its potential to help frontline experts retrieve information more efficiently.

Currently in the research and development phase and undergoing beta testing, the system uses vision-enabled large language models to generate responses based on retrieved source materials, with page-linked citations allowing users to trace information back to the original source. The underlying FME-approved source materials remain authoritative and are not replaced by AI-generated responses. 

The project is intended to investigate how conversational AI could help organize and navigate complex home hemodialysis information. Researchers are assessing its potential to improve knowledge accessibility and support clinical education workflows in the future. However, this technology is currently limited to the research and testing environment; it is not used to support patient care or clinical decision-making and does not replace existing approved materials or processes. 

Read the full abstract here: Improving confidence in home hemodialysis through conversational AI-driven knowledge access.

Watch the full Research in Brief below: