医学AI 月报
医学AI 月报
期间: 2026-06-01 ~ 2026-06-30 生成时间: 2026-07-20 17:36 论文总数: 23 | 行业新闻: 33 条
一、本期概览
📄 论文动态
本期论文以AI在医疗中的前沿应用为主,特别是大语言模型(LLM)的临床决策支持、检索增强生成(RAG)、自主AI代理、隐私保护及数据互操作性等方向,同时涵盖微生物组、眼科、急诊分诊、术后管理等具体场景,表明AI正从概念验证向临床落地快速推进。
🏭 行业动态
行业动态高度聚焦于AI医疗设备的监管与审批,包括FDA简化AI/ML设备审批流程、AHA致FDA信函、以及市场增长预测。同时Google DeepMind等巨头持续推动临床AI新模式,显示监管与技术创新双轮驱动行业成熟。
📊 高频期刊 TOP10
| 期刊 | IF | 篇数 |
|---|---|---|
| NPJ Digit Med | 12.4 | 2 |
| Nat Med | 58.7 | 2 |
| J Med Internet Res | 5.8 | 2 |
| Med | 8.6 | 2 |
| Digit Health | 23.8 | 2 |
| Gut | 19.8 | 1 |
| IEEE J Biomed Health Inform | 7.5 | 1 |
| Circ Res | 20.1 | 1 |
| Nature | 50.5 | 1 |
| Cancer Res | 10.0 | 1 |
二、研究主题聚类
大语言模型临床决策支持(7篇)
趋势: LLM被广泛用于急诊分诊、术后决策、血液肿瘤辅助决策、疫苗咨询等场景,结合RAG和记忆增强技术提升安全性和准确性,呈现多科室渗透趋势。
自主AI医疗代理与系统(3篇)
趋势: 从基础对话到自主代理(如EyeRAG、自主医学AI)成为热点,强调多轮交互、证据溯源和端到端能力,代表下一代AI医疗形态。
数据隐私与互操作性(3篇)
趋势: 针对医疗敏感数据,提出隐私预算LLM、数据互操作性重构方案,呼应实际部署中安全与兼容性核心痛点。
非AI基础研究(2篇)
趋势: CETP抑制剂、DGAT1抑制等药物靶点研究虽非AI,但为AI辅助药物发现提供上游生物学支撑,体现交叉价值。
三、重点推荐论文
Towards autonomous medical artificial intelligence agents.
Nature (IF:50.5) | PubMed ⚠️ | DOI
推荐理由: 发表于Nature,提出自主AI代理概念,被视为医疗AI未来范式的重要里程碑。
Clinical decision support in hematological malignancies using a case-grounded AI agent.
Nat Med (IF:58.7) | PubMed ⚠️ | DOI
推荐理由: Nat Med级别,首次在血液恶性肿瘤中验证案例驱动的AI辅助决策系统,临床转化意义重大。
EyeRAG: graph retrieval-augmented generation for safe and accurate clinical dialogue in ophthalmology.
NPJ Digit Med (IF:12.4) | PubMed ⚠️ | DOI
推荐理由: NPJ Digit Med,将图RAG引入眼科对话,提升安全性与准确率,为专科LLM应用树立典范。
Efficient Privacy-Budgeted Large Language Model for Sensitive Healthcare.
IEEE J Biomed Health Inform (IF:7.5) | PubMed ⚠️ | DOI
推荐理由: IEEE JBHI,解决医疗LLM隐私预算效率问题,是实际部署中的关键技术突破。
Rethinking healthcare data interoperability in the age of large language models.
Med (IF:8.6) | PubMed ⚠️ | DOI
推荐理由: Med期刊,重新定义LLM时代的数据互操作性,为AI与现有信息系统融合提供理论框架。
四、行业动态
How to Get your AI models FDA approved: AI/ML Regulatory Landscape | Encord
来源: | 链接
State of FDA Approval for AI algorithms
The number of AI and ML algorithms being approved by the US Food & Drug Administration (FDA) has accelerated dramatically in recent years.
950 AI or machin
FDA finalizes recommendations simplifying approval …
来源: | 链接 Dec 05, 2024 - 01:45 PM
FDA Update 900x400
The Food and Drug Administration yesterday released recommendations for streamlining the approval process for medical devices that use artificial intellige
AHA Letter to FDA on AI-enabled Medical Devices
来源: | 链接 Per section 201(h) of the Food, Drug and Cosmetic Act, certain AI applications are considered medical devices if they are “intended for use in the diagnosis of disease or other conditions, or in the c
Enabling a new model for healthcare with AI co-clinician
来源: | 链接 At Google DeepMind, our journey in medical AI has evolved from mastering examination-style tests of medical knowledge with MedPaLM, to matching physician performance in text-based simulated medical co
AI in Medicine: Exploring the Power of Google DeepMind
来源: | 链接 In this episode, discover how Google’s DeepMind AI is transforming medical care. We explore three key questions: How does DeepMind detect illnesses early? How will this technology improve healthcare f
五、展望
下阶段关注自主AI代理在真实临床环境中的部署验证、FDA对AI/ML设备持续更新监管框架、以及LLM结合隐私计算与联邦学习的合规实践。同时,RAG技术将更深度嵌入专科医疗系统,推动精准对话与决策支持。
由 Hermes Agent 自动生成