Area of Science:

• Artificial Intelligence in Medicine
• Clinical Informatics
• Healthcare Technology Integration

Background:

• Model Context Protocol (MCP) is an open standard for AI-tool connectivity.
• MCP was not initially designed for healthcare but offers potential for clinical decision support via natural language queries.
• Integrating AI tools into clinical workflows requires robust and reliable methods.

Purpose of the Study:

• To evaluate the feasibility and reliability of Model Context Protocol (MCP) for clinical decision support in pediatrics.
• To develop and validate a pediatric MCP server integrated with multiple clinical tools.
• To assess MCP's performance in processing real-world clinical case scenarios.

Main Methods:

• Development of a pediatric MCP server incorporating 46 distinct clinical tools.
• Testing the MCP server using a dataset of 32 diverse pediatric clinical cases.
• Evaluation of the system's accuracy in processing case information and retrieving relevant clinical data.

Main Results:

• The pediatric MCP server correctly processed 31 out of 32 tested clinical cases.
• Achieved a processing accuracy rate of 96.875% for the evaluated cases.
• Demonstrated successful integration and operation of multiple clinical tools via the MCP framework.

Conclusions:

• This study represents the first clinical validation of Model Context Protocol (MCP) technology.
• MCP technology shows high reliability and significant potential for clinical decision support applications in pediatrics.
• The findings support the broader adoption of MCP for enhancing AI-driven healthcare solutions.