Scholarlink

<p>This review examines the transformative potential of artificial intelligence (AI) agents that integrate cognitive reasoning with actionable decision-making in medical and clinical settings. By unifying deep learning, knowledge representation, and human–computer interaction, these agents enable context-aware diagnosis and personalized treatment planning. Case studies illustrate their success in disease diagnosis such as diabetic retinopathy and lung cancer, where AI agents have achieved notable diagnostic accuracy (e.g., 87% sensitivity for diabetic retinopathy screening and 98.7% for lung cancer detection). These agents also optimize treatment strategies, including personalized radiotherapy plans. However, challenges such as data privacy, interpretability, and real-world adaptability remain. We discuss existing limitations and propose future directions for developing autonomous and ethical AI agents in precision medicine, aiming to improve patient outcomes, reduce costs, and streamline clinical workflows.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Visual-action AI agents for medical diagnosis and treatment: advances and future outlook

  • 期刊论文

  • 发表日期：

    2025年9月25日

查看全文
  • Yuanqi Yao,
  • Yilun Luxue,
  • Xiang Chen,
  • Chengxing Shen,
  • Haodong Yang,
  • Seungmin Lee,
  • Tingli Chen,
  • Haiyan Ge,
  • Jie Shen

摘要

This review examines the transformative potential of artificial intelligence (AI) agents that integrate cognitive reasoning with actionable decision-making in medical and clinical settings. By unifying deep learning, knowledge representation, and human–computer interaction, these agents enable context-aware diagnosis and personalized treatment planning. Case studies illustrate their success in disease diagnosis such as diabetic retinopathy and lung cancer, where AI agents have achieved notable diagnostic accuracy (e.g., 87% sensitivity for diabetic retinopathy screening and 98.7% for lung cancer detection). These agents also optimize treatment strategies, including personalized radiotherapy plans. However, challenges such as data privacy, interpretability, and real-world adaptability remain. We discuss existing limitations and propose future directions for developing autonomous and ethical AI agents in precision medicine, aiming to improve patient outcomes, reduce costs, and streamline clinical workflows.