Model Context Protocol is Necessary for AI Agents in the Healthcare Domain

16. Tıp Bilişimi Kongresi, Ankara, Türkiye, 22 - 23 Mayıs 2025, (Yayınlanmadı)

The use of artificial intelligence (AI) in the medical domain

leads to a catastrophic revolution in the healthcare delivery system. The

modern healthcare system must deal with diverse data types such as

clinical notes, medical imaging, lab results, real-time monitoring, social

determinants of health, genetic data, medication history, etc. Most Clinical

Decision Support Systems (CDSS) still struggle with the complexity

of medical data collected for a patient. The future of CDSS is modular,

context-aware, and agent-driven. LLM-based AI agents designed to

understand complex medical situations and offer evidence-based recommendations

that ensure context-aware decision making for clinicians and

healthcare providers to improve clinical tasks. An AI agent goes beyond

just sending notifications; it acts as a reliable partner for healthcare

providers. Through AI Agents, personalized medicine became accessible

and affordable. However, data safety and ethical concerns are challenging

problems. The safe and reliable integration of LLM-based agents with

patient data is essential. Model Context Protocol (MCP) is a new standard

for two-way connection between AI agents, data sources and tools.

This study outlines a protocol tailored to the needs of the healthcare

domain, with a focus on applying the Model Context Protocol (MCP)

from a healthcare perspective.