AI
GraphDx: 逐次診断のためのコスト認識型知識強化マルチエージェントフレームワーク
GraphDx: A Cost-Aware Knowledge-Enhanced Multi-Agent Framework for Sequential Diagnosis
arxiv2026年7月20日
日本語要約
本論文は、知識グラフとマルチエージェントシステムをコスト認識と統合して逐次診断を強化する革新的なフレームワークGraphDxを紹介する。複雑な医療シナリオにおける診断精度と効率の向上を目指す。このアプローチは、よりインテリジェントで費用対効果の高いヘルスケアAIにつながる可能性がある。
English Summary
This paper introduces GraphDx, a novel framework that enhances sequential diagnosis by integrating knowledge graphs and multi-agent systems with cost awareness. It aims to improve diagnostic accuracy and efficiency in complex medical scenarios. This approach could lead to more intelligent and cost-effective healthcare AI.