MedLA: A Logic-Driven Multi-Agent Framework for Complex Medical Reasoning with Large Language Models

Researchers have developed MedLA, a new logic-driven multi-agent AI framework that uses large language models for complex medical reasoning. The system employs multiple AI agents that organize their reasoning into explicit logical trees and engage in structured discussions to resolve inconsistencies and reach consensus on medical questions.

• →MedLA uses syllogistic triads (major premise, minor premise, conclusion) to create transparent logical reasoning trees for medical decision-making.
• →The framework outperforms single-agent systems and static role-based approaches on challenging medical benchmarks like MedDDx.
• →Multiple AI agents engage in iterative discussions to detect and resolve logical inconsistencies in medical reasoning.
• →The system scales effectively across both open-source and commercial large language model backends.
• →MedLA achieves state-of-the-art performance while providing a generalizable paradigm for trustworthy medical AI reasoning.