Oracle Health推出与EHR集成的AI助手，将记录转化为通俗易懂的语言
Oracle Health Launches EHR-Integrated AI Assistant for Plain-Language Records
您需要了解的内容
AI驱动的 Oracle Health Patient Portal 现已在美国全境正式上线。
该解决方案可直接集成至 Oracle Health Electronic Health Record (EHR)，使患者能够在一个安全的环境中查看摘要、病史、实验室检查趋势和待办事项。
嵌入式、具备上下文感知能力的 AI 可将复杂的医疗文档、诊断结果、实验室检查结果和治疗方案转化为清晰易懂、便于理解的语言。
患者可以通过自然语言查询管理预约（例如请求进行背痛咨询），AI 会评估既往就诊记录，以推荐合适的临床医生和时间段。
该平台采用前沿 AI 模型构建，并配备严格的安全防护机制，禁止自主进行医学诊断，通过引用来源清晰标示 AI 生成的内容，并将所有数据保留在 Oracle 的安全临床环境中。
通俗易懂的 EHR 转译与上下文分析
传统患者门户通常会显示充满医学术语的原始临床记录，使医疗服务消费者感到困惑，并要求工作人员处理后续电话。Oracle Health Patient Portal 通过将 AI 模型直接置于患者经过验证的 EHR 记录基础上，解决了这一障碍：
基于病历的转译：
将诊断结果、当前用药方案、治疗方案和实验室检查结果转化为清晰、通俗易懂的摘要，无需患者将文档上传至外部工具。
上下文趋势分析：
患者可以就其病史提出对话式问题，例如
“帮我了解一下我这一年的胆固醇趋势”或“我的糖尿病管理得怎么样？”并获得直接根据其病历数据生成的回答。
自然语言预约：
处理对话式就诊请求（例如，“我想下周找我的医生进行背痛咨询”），评估既往与医疗服务提供者的关系，并在几个步骤内推荐合适的临床医生和开放时间段。
非诊断性安全防护机制：
包含严格的防护机制，禁止系统生成医疗建议或诊断结论。AI 生成的文本会以醒目方式标示并附带来源引用，所有个人健康信息均保存在 Oracle EHR 环境内，而不会导出至第三方模型。
Oracle Health 和 Life Sciences 执行副总裁兼总经理 Seema Verma 表示：“我们的 AI 驱动的 Oracle Health Patient Portal 从一开始就是为了帮助患者了解、管理和改善自身健康而构建的。通过让患者能够轻松管理医疗服务并进行数字化沟通，我们也在帮助医疗团队减少行政负担，将时间集中于最重要的事情：提供尽可能优质的医疗服务。”
What You Should Know The AI-powered Oracle Health Patient Portal is now generally available across the United States.
The solution integrates directly into the Oracle Health
Electronic Health Record (EHR), allowing patients to view summaries, medical histories, lab trends, and pending tasks in one secure environment.
Embedded, context-aware AI translates complex medical documentation, diagnoses, lab results, and treatment options into plain, accessible language.
Patients can manage scheduling via natural language queries (e.g., requesting a back pain consultation), with the AI evaluating prior visits to recommend appropriate clinicians and time slots.
Built using frontier AI models with strict safety guardrails, the platform prohibits autonomous medical diagnoses, clearly highlights AI-generated content with citations, and keeps all data within Oracle’s secure clinical environment.
Plain-Language EHR Translation and Contextual Analytics
Legacy patient portals often display raw clinical records filled with medical jargon, leaving health consumers confused and requiring staff to handle follow-up phone calls. The Oracle Health Patient Portal addresses this friction by grounding AI models directly in the patient’s verified EHR record:
Chart-Grounded Translation:
Translates diagnoses, active medication regimens, treatment options, and lab results into clear, plain-language summaries without requiring patients to upload documents to external tools.
Contextual Trend Analysis:
Allows patients to ask conversational questions about their medical history—such as “Help me understand my cholesterol trends over the year” or “How am I managing my diabetes?” —and receive answers drawn directly from their chart data.
Natural Language Scheduling:
Processes conversational visit requests (e.g., “I want to see my doctor for a back pain consultation next week” ), evaluates historical provider relationships, and recommends appropriate clinicians and open time slots in a few steps.
Non-Diagnostic Safety Guardrails:
Includes strict guardrails that prohibit the system from generating medical advice or diagnostic conclusions. AI-generated text is visually highlighted with source citations, and all personal health information remains secured inside the Oracle EHR environment rather than exported to third-party models.
“Our AI-powered Oracle Health Patient Portal is built from the ground up to help patients understand, manage, and improve their health,” said Seema Verma, Executive Vice President and General Manager of Oracle Health and Life Sciences. “By making it easy for patients to navigate their care and communicate digitally, we’re also helping care teams reduce administrative burden and focus their time on what matters most: delivering the best possible care.” Reader Interactions